blender/source/gameengine/VideoTexture/ImageBase.h

357 lines
9.5 KiB
C
Raw Normal View History

VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
/* $Id$
-----------------------------------------------------------------------------
This source file is part of blendTex library
Copyright (c) 2007 The Zdeno Ash Miklas
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU Lesser General Public License as published by the Free Software
Foundation; either version 2 of the License, or (at your option) any later
version.
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public License along with
this program; if not, write to the Free Software Foundation, Inc., 59 Temple
Place - Suite 330, Boston, MA 02111-1307, USA, or go to
http://www.gnu.org/copyleft/lesser.txt.
-----------------------------------------------------------------------------
*/
#if !defined IMAGEBASE_H
#define IMAGEBASE_H
#include "Common.h"
#include <vector>
#include <PyObjectPlus.h>
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
#include "PyTypeList.h"
#include "FilterBase.h"
// forward declarations
struct PyImage;
class ImageSource;
/// type for list of image sources
typedef std::vector<ImageSource*> ImageSourceList;
/// base class for image filters
class ImageBase
{
public:
/// constructor
ImageBase (bool staticSrc = false);
/// destructor
virtual ~ImageBase (void);
/// release contained objects, if returns true, object should be deleted
virtual bool release (void);
VideoTexture: improvements to image data access API. - Use BGL buffer instead of string for image data. - Add buffer interface to image source. - Allow customization of pixel format. - Add valid property to check if the image data is available. The image property of all Image source objects will now return a BGL 'buffer' object. Previously it was returning a string, which was not working at all with Python 3.1. The BGL buffer type allows sequence access to bytes and is directly usable in BGL OpenGL wrapper functions. The buffer is formated as a 1 dimensional array of bytes with 4 bytes per pixel in RGBA order. BGL buffers will also be accepted in the ImageBuff load() and plot() functions. It is possible to customize the pixel format by using the VideoTexture.imageToArray(image, mode) function: the first argument is a Image source object, the second optional argument is a format string using the R, G, B, A, 0 and 1 characters. For example "BGR" means that each pixel will be 3 bytes, corresponding to the Blue, Green and Red channel in that order. Use 0 for a fixed hex 00 value, 1 for hex FF. The default mode is "RGBA". All Image source objects now support the buffer interface which allows to create memoryview objects for direct access to the image internal buffer without memory copy. The buffer format is one dimensional array of bytes with 4 bytes per pixel in RGBA order. The buffer is writable, which allows custom modifications of the image data. v = memoryview(source) A bug in the Python 3.1 buffer API will cause a crash if the memoryview object cannot be created. Therefore, you must always check first that an image data is available before creating a memoryview object. Use the new valid attribute for that: if source.valid: v = memoryview(source) ... Note: the BGL buffer object itself does not yet support the buffer interface. Note: the valid attribute makes sense only if you use image source in conjunction with texture object like this: # refresh texture but keep image data in memory texture.refresh(False) if texture.source.valid: v = memoryview(texture.source) # process image ... # invalidate image for next texture refresh texture.source.refresh() Limitation: While memoryview objects exist, the image cannot be resized. Resizing occurs with ImageViewport objects when the viewport size is changed or with ImageFFmpeg when a new image is reloaded for example. Any attempt to resize will cause a runtime error. Delete the memoryview objects is you want to resize an image source object.
2010-02-21 22:20:00 +00:00
/// is an image available
bool isImageAvailable(void)
{ return m_avail; }
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
/// get image
unsigned int * getImage (unsigned int texId = 0, double timestamp=-1.0);
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
/// get image size
short * getSize (void) { return m_size; }
/// get image buffer size
unsigned long getBuffSize (void)
{ return m_size[0] * m_size[1] * sizeof(unsigned int); }
/// refresh image - invalidate its current content
virtual void refresh (void);
/// get scale
bool getScale (void) { return m_scale; }
/// set scale
void setScale (bool scale) { m_scale = scale; m_scaleChange = true; }
/// get vertical flip
bool getFlip (void) { return m_flip; }
/// set vertical flip
void setFlip (bool flip) { m_flip = flip; }
/// get source object
PyImage * getSource (const char * id);
/// set source object, return true, if source was set
bool setSource (const char * id, PyImage * source);
/// get pixel filter
PyFilter * getFilter (void) { return m_pyfilter; }
/// set pixel filter
void setFilter (PyFilter * filt);
/// calculate size (nearest power of 2)
static short calcSize (short size);
VideoTexture: improvements to image data access API. - Use BGL buffer instead of string for image data. - Add buffer interface to image source. - Allow customization of pixel format. - Add valid property to check if the image data is available. The image property of all Image source objects will now return a BGL 'buffer' object. Previously it was returning a string, which was not working at all with Python 3.1. The BGL buffer type allows sequence access to bytes and is directly usable in BGL OpenGL wrapper functions. The buffer is formated as a 1 dimensional array of bytes with 4 bytes per pixel in RGBA order. BGL buffers will also be accepted in the ImageBuff load() and plot() functions. It is possible to customize the pixel format by using the VideoTexture.imageToArray(image, mode) function: the first argument is a Image source object, the second optional argument is a format string using the R, G, B, A, 0 and 1 characters. For example "BGR" means that each pixel will be 3 bytes, corresponding to the Blue, Green and Red channel in that order. Use 0 for a fixed hex 00 value, 1 for hex FF. The default mode is "RGBA". All Image source objects now support the buffer interface which allows to create memoryview objects for direct access to the image internal buffer without memory copy. The buffer format is one dimensional array of bytes with 4 bytes per pixel in RGBA order. The buffer is writable, which allows custom modifications of the image data. v = memoryview(source) A bug in the Python 3.1 buffer API will cause a crash if the memoryview object cannot be created. Therefore, you must always check first that an image data is available before creating a memoryview object. Use the new valid attribute for that: if source.valid: v = memoryview(source) ... Note: the BGL buffer object itself does not yet support the buffer interface. Note: the valid attribute makes sense only if you use image source in conjunction with texture object like this: # refresh texture but keep image data in memory texture.refresh(False) if texture.source.valid: v = memoryview(texture.source) # process image ... # invalidate image for next texture refresh texture.source.refresh() Limitation: While memoryview objects exist, the image cannot be resized. Resizing occurs with ImageViewport objects when the viewport size is changed or with ImageFFmpeg when a new image is reloaded for example. Any attempt to resize will cause a runtime error. Delete the memoryview objects is you want to resize an image source object.
2010-02-21 22:20:00 +00:00
/// number of buffer pointing to m_image, public because not handled by this class
int m_exports;
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
protected:
/// image buffer
unsigned int * m_image;
/// image buffer size
unsigned int m_imgSize;
/// image size
short m_size[2];
/// image is available
bool m_avail;
/// scale image to power 2 sizes
bool m_scale;
/// scale was changed
bool m_scaleChange;
/// flip image vertically
bool m_flip;
/// source image list
ImageSourceList m_sources;
/// flag for disabling addition and deletion of sources
bool m_staticSources;
/// pixel filter
PyFilter * m_pyfilter;
/// initialize image data
void init (short width, short height);
/// find source
ImageSourceList::iterator findSource (const char * id);
/// create new source
virtual ImageSource * newSource (const char * id) { return NULL; }
/// check source sizes
bool checkSourceSizes (void);
/// calculate image from sources and set its availability
virtual void calcImage (unsigned int texId, double ts) {}
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
/// perform loop detection
bool loopDetect (ImageBase * img);
/// template for image conversion
template<class FLT, class SRC> void convImage (FLT & filter, SRC srcBuff,
short * srcSize)
{
// destination buffer
unsigned int * dstBuff = m_image;
// pixel size from filter
unsigned int pixSize = filter.firstPixelSize();
// if no scaling is needed
if (srcSize[0] == m_size[0] && srcSize[1] == m_size[1])
// if flipping isn't required
if (!m_flip)
// copy bitmap
for (short y = 0; y < m_size[1]; ++y)
for (short x = 0; x < m_size[0]; ++x, ++dstBuff, srcBuff += pixSize)
// copy pixel
*dstBuff = filter.convert(srcBuff, x, y, srcSize, pixSize);
// otherwise flip image top to bottom
else
{
// go to last row of image
srcBuff += srcSize[0] * (srcSize[1] - 1) * pixSize;
// copy bitmap
for (short y = m_size[1] - 1; y >= 0; --y, srcBuff -= 2 * srcSize[0] * pixSize)
for (short x = 0; x < m_size[0]; ++x, ++dstBuff, srcBuff += pixSize)
// copy pixel
*dstBuff = filter.convert(srcBuff, x, y, srcSize, pixSize);
}
// else scale picture (nearest neighbour)
else
{
// interpolation accumulator
int accHeight = srcSize[1] >> 1;
// if flipping is required
if (m_flip)
// go to last row of image
srcBuff += srcSize[0] * (srcSize[1] - 1) * pixSize;
// process image rows
for (int y = 0; y < srcSize[1]; ++y)
{
// increase height accum
accHeight += m_size[1];
// if pixel row has to be drawn
if (accHeight >= srcSize[1])
{
// decrease accum
accHeight -= srcSize[1];
// width accum
int accWidth = srcSize[0] >> 1;
// process row
for (int x = 0; x < srcSize[0]; ++x)
{
// increase width accum
accWidth += m_size[0];
// if pixel has to be drawn
if (accWidth >= srcSize[0])
{
// decrease accum
accWidth -= srcSize[0];
// convert pixel
*dstBuff = filter.convert(srcBuff, x, m_flip ? srcSize[1] - y - 1 : y,
srcSize, pixSize);
// next pixel
++dstBuff;
}
// shift source pointer
srcBuff += pixSize;
}
}
// if pixel row will not be drawn
else
// move source pointer to next row
srcBuff += pixSize * srcSize[0];
// if y flipping is required
if (m_flip)
// go to previous row of image
srcBuff -= 2 * pixSize * srcSize[0];
}
}
}
// template for specific filter preprocessing
template <class F, class SRC> void filterImage (F & filt, SRC srcBuff, short * srcSize)
{
// find first filter in chain
FilterBase * firstFilter = NULL;
if (m_pyfilter != NULL) firstFilter = m_pyfilter->m_filter->findFirst();
// if first filter is available
if (firstFilter != NULL)
{
// python wrapper for filter
PyFilter pyFilt;
pyFilt.m_filter = &filt;
// set specified filter as first in chain
firstFilter->setPrevious(&pyFilt, false);
// convert video image
convImage(*(m_pyfilter->m_filter), srcBuff, srcSize);
// delete added filter
firstFilter->setPrevious(NULL, false);
}
// otherwise use given filter for conversion
else convImage(filt, srcBuff, srcSize);
// source was processed
m_avail = true;
}
};
// python structure for image filter
struct PyImage
{
PyObject_HEAD
// source object
ImageBase * m_image;
};
// size of id
const int SourceIdSize = 32;
/// class for source of image
class ImageSource
{
public:
/// constructor
ImageSource (const char * id);
/// destructor
virtual ~ImageSource (void);
/// get id
const char * getId (void) { return m_id; }
/// compare id to argument
bool is (const char * id);
/// get source object
PyImage * getSource (void) { return m_source; }
/// set source object
void setSource (PyImage * source);
/// get image from source
unsigned int * getImage (double ts=-1.0);
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
/// get buffered image
unsigned int * getImageBuf (void) { return m_image; }
/// refresh source
void refresh (void);
/// get image size
short * getSize (void)
{
static short defSize [] = {0, 0};
return m_source != NULL ? m_source->m_image->getSize() : defSize;
}
protected:
/// id of source
char m_id [SourceIdSize];
/// pointer to source structure
PyImage * m_source;
/// buffered image from source
unsigned int * m_image;
private:
/// default constructor is forbidden
ImageSource (void) {}
};
// list of python image types
extern PyTypeList pyImageTypes;
// functions for python interface
// object initialization
template <class T> static int Image_init (PyObject * pySelf, PyObject * args, PyObject * kwds)
{
PyImage * self = reinterpret_cast<PyImage*>(pySelf);
// create source object
if (self->m_image != NULL) delete self->m_image;
self->m_image = new T();
// initialization succeded
return 0;
}
// object allocation
PyObject * Image_allocNew (PyTypeObject * type, PyObject * args, PyObject * kwds);
// object deallocation
void Image_dealloc (PyImage * self);
// get image data
VideoTexture: improvements to image data access API. - Use BGL buffer instead of string for image data. - Add buffer interface to image source. - Allow customization of pixel format. - Add valid property to check if the image data is available. The image property of all Image source objects will now return a BGL 'buffer' object. Previously it was returning a string, which was not working at all with Python 3.1. The BGL buffer type allows sequence access to bytes and is directly usable in BGL OpenGL wrapper functions. The buffer is formated as a 1 dimensional array of bytes with 4 bytes per pixel in RGBA order. BGL buffers will also be accepted in the ImageBuff load() and plot() functions. It is possible to customize the pixel format by using the VideoTexture.imageToArray(image, mode) function: the first argument is a Image source object, the second optional argument is a format string using the R, G, B, A, 0 and 1 characters. For example "BGR" means that each pixel will be 3 bytes, corresponding to the Blue, Green and Red channel in that order. Use 0 for a fixed hex 00 value, 1 for hex FF. The default mode is "RGBA". All Image source objects now support the buffer interface which allows to create memoryview objects for direct access to the image internal buffer without memory copy. The buffer format is one dimensional array of bytes with 4 bytes per pixel in RGBA order. The buffer is writable, which allows custom modifications of the image data. v = memoryview(source) A bug in the Python 3.1 buffer API will cause a crash if the memoryview object cannot be created. Therefore, you must always check first that an image data is available before creating a memoryview object. Use the new valid attribute for that: if source.valid: v = memoryview(source) ... Note: the BGL buffer object itself does not yet support the buffer interface. Note: the valid attribute makes sense only if you use image source in conjunction with texture object like this: # refresh texture but keep image data in memory texture.refresh(False) if texture.source.valid: v = memoryview(texture.source) # process image ... # invalidate image for next texture refresh texture.source.refresh() Limitation: While memoryview objects exist, the image cannot be resized. Resizing occurs with ImageViewport objects when the viewport size is changed or with ImageFFmpeg when a new image is reloaded for example. Any attempt to resize will cause a runtime error. Delete the memoryview objects is you want to resize an image source object.
2010-02-21 22:20:00 +00:00
PyObject * Image_getImage (PyImage * self, char * mode);
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
// get image size
PyObject * Image_getSize (PyImage * self, void * closure);
// refresh image - invalidate current content
PyObject * Image_refresh (PyImage * self);
// get scale
PyObject * Image_getScale (PyImage * self, void * closure);
// set scale
int Image_setScale (PyImage * self, PyObject * value, void * closure);
// get flip
PyObject * Image_getFlip (PyImage * self, void * closure);
// set flip
int Image_setFlip (PyImage * self, PyObject * value, void * closure);
// get filter source object
PyObject * Image_getSource (PyImage * self, PyObject * args);
// set filter source object
PyObject * Image_setSource (PyImage * self, PyObject * args);
// get pixel filter object
PyObject * Image_getFilter (PyImage * self, void * closure);
// set pixel filter object
int Image_setFilter (PyImage * self, PyObject * value, void * closure);
VideoTexture: improvements to image data access API. - Use BGL buffer instead of string for image data. - Add buffer interface to image source. - Allow customization of pixel format. - Add valid property to check if the image data is available. The image property of all Image source objects will now return a BGL 'buffer' object. Previously it was returning a string, which was not working at all with Python 3.1. The BGL buffer type allows sequence access to bytes and is directly usable in BGL OpenGL wrapper functions. The buffer is formated as a 1 dimensional array of bytes with 4 bytes per pixel in RGBA order. BGL buffers will also be accepted in the ImageBuff load() and plot() functions. It is possible to customize the pixel format by using the VideoTexture.imageToArray(image, mode) function: the first argument is a Image source object, the second optional argument is a format string using the R, G, B, A, 0 and 1 characters. For example "BGR" means that each pixel will be 3 bytes, corresponding to the Blue, Green and Red channel in that order. Use 0 for a fixed hex 00 value, 1 for hex FF. The default mode is "RGBA". All Image source objects now support the buffer interface which allows to create memoryview objects for direct access to the image internal buffer without memory copy. The buffer format is one dimensional array of bytes with 4 bytes per pixel in RGBA order. The buffer is writable, which allows custom modifications of the image data. v = memoryview(source) A bug in the Python 3.1 buffer API will cause a crash if the memoryview object cannot be created. Therefore, you must always check first that an image data is available before creating a memoryview object. Use the new valid attribute for that: if source.valid: v = memoryview(source) ... Note: the BGL buffer object itself does not yet support the buffer interface. Note: the valid attribute makes sense only if you use image source in conjunction with texture object like this: # refresh texture but keep image data in memory texture.refresh(False) if texture.source.valid: v = memoryview(texture.source) # process image ... # invalidate image for next texture refresh texture.source.refresh() Limitation: While memoryview objects exist, the image cannot be resized. Resizing occurs with ImageViewport objects when the viewport size is changed or with ImageFFmpeg when a new image is reloaded for example. Any attempt to resize will cause a runtime error. Delete the memoryview objects is you want to resize an image source object.
2010-02-21 22:20:00 +00:00
// check if a buffer can be extracted
PyObject * Image_valid(PyImage * self, void * closure);
// for buffer access to PyImage objects
extern PyBufferProcs imageBufferProcs;
VideoTexture module. The only compilation system that works for sure is the MSVC project files. I've tried my best to update the other compilation system but I count on the community to check and fix them. This is Zdeno Miklas video texture plugin ported to trunk. The original plugin API is maintained (can be found here http://home.scarlet.be/~tsi46445/blender/blendVideoTex.html) EXCEPT for the following: The module name is changed to VideoTexture (instead of blendVideoTex). A new (and only) video source is now available: VideoFFmpeg() You must pass 1 to 4 arguments when you create it (you can use named arguments): VideoFFmpeg(file) : play a video file VideoFFmpeg(file, capture, rate, width, height) : start a live video capture file: In the first form, file is a video file name, relative to startup directory. It can also be a URL, FFmpeg will happily stream a video from a network source. In the second form, file is empty or is a hint for the format of the video capture. In Windows, file is ignored and should be empty or not specified. In Linux, ffmpeg supports two types of device: VideoForLinux and DV1394. The user specifies the type of device with the file parameter: [<device_type>][:<standard>] <device_type> : 'v4l' for VideoForLinux, 'dv1394' for DV1394; default to 'v4l' <standard> : 'pal', 'secam' or 'ntsc', default to 'ntsc' The driver name is constructed automatically from the device types: v4l : /dev/video<capture> dv1394: /dev/dv1394/<capture> If you have different driver name, you can specify the driver name explicitely instead of device type. Examples of valid file parameter: /dev/v4l/video0:pal /dev/ieee1394/1:ntsc dv1394:ntsc v4l:pal :secam capture: Defines the index number of the capture source, starting from 0. The first capture device is always 0. The VideoTexutre modules knows that you want to start a live video capture when you set this parameter to a number >= 0. Setting this parameter < 0 indicates a video file playback. Default value is -1. rate: the capture frame rate, by default 25 frames/sec width: height: Width and height of the video capture in pixel, default value 0. In Windows you must specify these values and they must fit with the capture device capability. For example, if you have a webcam that can capture at 160x120, 320x240 or 640x480, you must specify one of these couple of values or the opening of the video source will fail. In Linux, default values are provided by the VideoForLinux driver if you don't specify width and height. Simple example ************** 1. Texture definition script: import VideoTexture contr = GameLogic.getCurrentController() obj = contr.getOwner() if not hasattr(GameLogic, 'video'): matID = VideoTexture.materialID(obj, 'MAVideoMat') GameLogic.video = VideoTexture.Texture(obj, matID) GameLogic.vidSrc = VideoTexture.VideoFFmpeg('trailer_400p.ogg') # Streaming is also possible: #GameLogic.vidSrc = VideoTexture.VideoFFmpeg('http://10.32.1.10/trailer_400p.ogg') GameLogic.vidSrc.repeat = -1 # If the video dimensions are not a power of 2, scaling must be done before # sending the texture to the GPU. This is done by default with gluScaleImage() # but you can also use a faster, but less precise, scaling by setting scale # to True. Best approach is to convert the video offline and set the dimensions right. GameLogic.vidSrc.scale = True # FFmpeg always delivers the video image upside down, so flipping is enabled automatically #GameLogic.vidSrc.flip = True if contr.getSensors()[0].isPositive(): GameLogic.video.source = GameLogic.vidSrc GameLogic.vidSrc.play() 2. Texture refresh script: obj = GameLogic.getCurrentController().getOwner() if hasattr(GameLogic, 'video') != 0: GameLogic.video.refresh(True) You can download this demo here: http://home.scarlet.be/~tsi46445/blender/VideoTextureDemo.blend http://home.scarlet.be/~tsi46445/blender/trailer_400p.ogg
2008-10-31 22:35:52 +00:00
#endif