一种基于交叉视觉皮质模型的图像编码算法

AN ALGORITHM OF IMAGE ENCODING BASED ON ICM

根据交叉视觉皮质模型(ICM)具有的图像分块功能对图像分块。利用已有的寻找拐点(corner点)方法寻找分块后的corner点,接着对corner点所影响的区域进行无损编码,对不受影响的点进行有损编码(DCT编码)。实验证明,这种编码在保证一定压缩率的情况下可以清晰地显示图像的关键点(也即corner点)所影响的区域。

The image was segmented into blocks according to the function of image segmentation which ICM（intersecting cortical model） has.An existing way of finding the corners is employed to seek corners after segmentation has been done,and then follows with lossless encoding on the blocks of the image which are affected by the corners;for other unaffected blocks,lossy encoding（DCT encoding） is exerted on them.Experimental results illustrate that this encoding method can clearly display the areas affected by key points（i.e.comers） of the image in condition of ensuring certain compression rate.