zoukankan      html  css  js  c++  java
  • Faster Alternatives to glReadPixels and glTexImage2D in OpenGL ES

    In the development of Shou, I’ve been using GLSL with NEON to manipulate image rotation, scaling and color conversion, before send them to video encoder.

    So I need a very efficient way to transfer pixels between OpenGL and memory space. TheglTexImage2D and glReadPixels performance are very unacceptable, especially for some specific vendors, e.g. Samsung Galaxy devices with Exynos chip.

    Compared to glTex(Sub)Image2D, the glReadPixels is the real bottleneck, which blocks all OpenGL pipeline and results in about 100ms delay for a standard 720P frame read back.

    Here I will share two standard OpenGL approaches to achieve really faster pixels pack, which should be available on all OpenGL implementations. Only glReadPixels will be discussed, as the glTexImage2D should have the same usage.

    Pixel Buffer Object

    PBO is not introduced until OpenGL ES 3.0, which is available since Android 4.3. The pixels pack operation will be reduced to about 5ms using PBO.

    PBO is created just like any other buffer objects:

    glGenBuffers(1, &pbo_id); glBindBuffer(GL_PIXEL_PACK_BUFFER, pbo_id); glBufferData(GL_PIXEL_PACK_BUFFER, pbo_size, 0, GL_DYNAMIC_READ); glBindBuffer(GL_PIXEL_PACK_BUFFER, 0);

    According to the reference of glReadPixels:

    If a non-zero named buffer object is bound to the GL_PIXEL_PACK_BUFFER target (see glBindBuffer) while a block of pixels is requested, data is treated as a byte offset into the buffer object’s data store rather than a pointer to client memory.

    When we need to read pixels from an FBO:

    glReadBuffer(GL_COLOR_ATTACHMENT0); glBindBuffer(GL_PIXEL_PACK_BUFFER, pbo_id); glReadPixels(0, 0, width, height, GL_RGBA, GL_UNSIGNED_BYTE, 0); GLubyte *ptr = glMapBufferRange(GL_PIXEL_PACK_BUFFER, 0, pbo_size, GL_MAP_READ_BIT); memcpy(pixels, ptr, pbo_size); glUnmapBuffer(GL_PIXEL_PACK_BUFFER); glBindBuffer(GL_PIXEL_PACK_BUFFER, 0);

    In a real project, we may consider using double or triple PBOs to improve the performance.

    EGLImage

    EGL_KHR_image_base is a completed EGL extension, which achieves the same performance as PBO, but only require OpenGL-ES 1.1 or 2.0.

    The function to create an EGLImageKHR is

    EGLImageKHR eglCreateImageKHR(EGLDisplay dpy,                               EGLContext ctx,                               EGLenum target,                               EGLClientBuffer buffer,                               const EGLint *attrib_list)

    The Android EGL implementation frameworks/native/opengl/libagl/egl.cpp implies that theEGLDisplay should be a valid display, the EGLClientBuffer type should be ANativeWindowBuffer, the EGLContext can only be EGL_NO_CONTEXT, and the target can only beEGL_NATIVE_BUFFER_ANDROID.

    All the other parameters are obvious, except for the ANativeWindowBuffer, which is defined insystem/core/include/system/window.h.

    To allocate an ANativeWindowBuffer, Android has a simple wrapper called GraphicBuffer, defined in frameworks/native/include/ui/GraphicBuffer.h.

    GraphicBuffer *window = new GraphicBuffer(width, height, PIXEL_FORMAT_RGBA_8888, GraphicBuffer::USAGE_SW_READ_OFTEN | GraphicBuffer::USAGE_HW_TEXTURE);  struct ANativeWindowBuffer *buffer = window->getNativeBuffer(); EGLImageKHR *image = eglCreateImageKHR(eglGetCurrentDisplay(), EGL_NO_CONTEXT, EGL_NATIVE_BUFFER_ANDROID, *attribs);

    Then anytime we want to read pixels from an FBO, we should use one of the two methods below:

    void EGLImageTargetTexture2DOES(enum target, eglImageOES image)  void EGLImageTargetRenderbufferStorageOES(enum target, eglImageOES image)

    These two methods will establishes all the properties of the target GL_TEXTURE_2D orGL_RENDERBUFFER.

    uint8_t *ptr; glBindTexture(GL_TEXTURE_2D, texture_id); glEGLImageTargetTexture2DOES(GL_TEXTURE_2D, image);  window->lock(GraphicBuffer::USAGE_SW_READ_OFTEN, &ptr); memcpy(pixels, ptr, width * height * 4); window->unlock();

    References

    1. GL_PIXEL_PACK_BUFFERhttp://www.khronos.org/opengles/sdk/docs/man3/xhtml/glMapBufferRange.xml
    2. EGL_KHR_image_basehttp://www.khronos.org/registry/egl/extensions/KHR/EGL_KHR_image_base.txt
    3. GL_OES_EGL_imagehttp://www.khronos.org/registry/gles/extensions/OES/OES_EGL_image.txt
    4. Using direct textures on Android http://snorp.net/2011/12/16/android-direct-texture.html
    5. Using OpenGL ES to Accelerate Apps with Legacy 2D GUIs http://software.intel.com/en-us/articles/using-opengl-es-to-accelerate-apps-with-legacy-2d-guis
    6. iOS solution http://stackoverflow.com/questions/9550297/faster-alternative-to-glreadpixels-in-iphone-opengl-es-2-0
  • 相关阅读:
    【如何入门ACM】
    HDU
    HDU 6107 Typesetting
    bzoj 3223: Tyvj 1729 文艺平衡树
    51Nod 1781 跑的比谁都快
    51Nod 1331 狭窄的通道
    51Nod 1555 布丁怪
    hihocoder 1035 : 自驾旅行 III
    51Nod 1196 字符串的数量
    51Nod 1530 稳定方块
  • 原文地址:https://www.cnblogs.com/nsnow/p/3863792.html
Copyright © 2011-2022 走看看