The invention discloses a high-capacity image steganography method. The method comprises the following steps: carrying out scrambling processing on a carrier image; Dividing the scrambled image into N* N sub-blocks and carrying out DCT (Discrete Cosine Transform); Classifying each DCT block, and selecting an embedded block; Element values at selected positions are extracted from a matrix obtainedafter the DCT coefficient of each embedded block is scanned in a Z shape, and a coefficient matrix is obtained; Extracting element values which are the same as the element values at the selected positions from a matrix obtained after Z-shaped scanning of the Q table to form a quantization matrix; Calculating a division matrix and block capacity through the quantization matrix; Dividing the bit stream of the secret information according to the block capacity; Calculating an embedded matrix for the divided bit stream; Calculating a new coefficient matrix; Replacing the DCT coefficient with theelement value in the new coefficient matrix, and carrying out inverse operation of Z-shaped scanning to obtain a new DCT coefficient; And carrying out DCT inverse transformation on each sub-block andthen carrying out inverse scrambling processing. The invention also discloses a secret message extraction method. And the image complexity is high and the embedding capacity is large.