彩色身份证件照片的混合压缩算法

Hybrid Compression Algorithm for Color Personal Identification Pictures

提出了一个彩色身份证件照片的混合压缩算法 .该算法把 FC2 VQ(Feature Correction tow- stage VectorQuantization)算法和著名的 EZW类算法结合在一起 ,充分利用这两类算法的优点对彩色身份证件照片进行压缩 ,得到一幅视觉效果最好的压缩照片 .在个人身份证件实验系统中 ,用这种算法将一幅 96× 12 8的 2 4- b彩色照片连同两枚指纹的特征数据及必要的个人信息存储在一个 PDF417二维条形码 (存储空间为 10 2 4个字节 )中 ,照片的重建质量可以接受 .

This paper presents a hybrid coding algorithm for color photo ID pictures. A novel technique for image coding by taking advantage of the FC2VQ algorithm and the EZW algorithm is addressed. A region of facial features (ROFF) is detected and compressed as accurately as possible. Therefore, the hybrid method gives the 'best visual' quality of the compressed color image since the storage of media is fully used for compression of the ROFF. It was used in experimental digital personal fingerprint photo ID system, which was needed to store two fingerprints feature data and a color photo ID picture with other personal information in a PDF 417 barcode (1024 bytes). The quality of the compressed 96×128 color photo of ID pictures is satisfied.