使用二维离散余弦变换(DCT)进行图象压缩。原始图象被转换成为8*8大小的块进行压缩和重构。 Cnum:为压缩时保留的系数的个数 MSM:均方误差压缩应该做到在最合理的近似原图象的情况下使用最少的系数。函数：dctmtx:返回一个n*n阶DCT变换 blkproc:对图象进行不同的块处理。 im2col:重排图象块为矩阵列。注意：只适用于灰度图象。-The use of two-dimensional discrete cosine transform (DCT) for image compression. Original image is converted to become the size of 8* 8 block compression and reconstruction. Cnum: reservations for the compression factor of the number of MSM: mean square error compression should be done in the most reasonable approximation of the original image using the least coefficient. Function: dctmtx: returns an n* n-order DCT transform blkproc: image processing of different blocks. im2col: rearrangement of the matrix block out images. Note: only applicable to grayscale images.