摘要
为了实现对数字图像有效的传输、保存、检索、加密、恢复。经常对图像进行DCT变换,在分析DCT变换直流系数的特点和性质基础上,设计了一种用DCT直流系数生成原图像轮廓的算法,将原图像进行互不重叠的分块,对每个块进行DCT变换,比较各块直流系数与直流系数平均值的大小得到二值矩阵,用该二值矩阵构造的二值图像即为原始图像轮廓。如果在图像处理过程中对图像进行了分块DCT变换,其他数据丢失,只保留了各块的直流系数,可以生成原图像的轮廓,也可以生成原图像的缩小图像。证明了DC系数二值化与平均像素二值化生成的原图像轮廓是一样的。
In order to achieve effective the digital image transmission, storage, retrieval, encryption, recovery. Often take image DCT transformation, analyze characteristics and nature of the DCT-DC coefficient, Designed an algorithm base on DCT-DC coefficient generated the contour of original image. The original images are not overlapping block of each block to transform DCT, get a binary ma- trix by comparing DC coefficient of each block and DC coefficient average. Using the matrix generation binary image were contour of the original image. If taking images DCT transformation when image processing, other data loss, retaining only the DC coefficient, Can generate the outline and reduce the image of the original image. Proof of DC coefficient binary operation and the average value of pixels binary operation generation outline of the image are the same.
出处
《微计算机信息》
2009年第25期121-122,115,共3页
Control & Automation
基金
基金申请人:吴柯
项目名称:图像变换域数字水印算法与评估方法研究
基金颁发部门:湖南省教育厅(08C389)
