| 摘要 |
Techniques are disclosed for compressing image data using the parallel instructions available on a vector processor. In an embodiment, a source image in RGBA format is received. A series of vector processor instructions are used to perform parallel processing steps of splitting the image into separate R, G, B, and A channels, then converting it into a YCoCg colorspace, compressing the Y, Co, and Cg channels separately, then saving an output image corresponding to a compressed version of the source image. |
