一种改进的二值图像矩形划分编码

摘要

The Rectangular Partition Coding ( RPC ), according to the principle that two points can determine a rectangle, divides binary image into rectangular areas so as to achieve the goal of image data compression. Since the black and white pixels are mixed in edge region of the image,when dividing the images this will lead to many long narrow rectangles being marked,the compression results obtained are unsatisfactory with low compression ratio. In response to this limitation,the paper proposes the rectangular window search algorithm as an improvement which divides all the black and white pixel areas in the image with rectangular window sliding search to obtain fewer rectangles that can cover the pixels area. Experiment proves that the improved algorithm can greatly reduce the number of the rectangle and raise the compression ratio.%矩形划分编码根据两点能够确定一个矩形的原理,把二值图像划分为矩形区域从而达到压缩图像数据的目的.由于图像边缘区域黑白像素混杂,导致图像划分时会标识出许多长而狭窄的矩形,这样得到的压缩效果并不理想,压缩比不高.针对这一局限进行了改进,提出了矩形窗口搜索算法,对图像中黑白像素区域都进行矩形窗口滑动搜索划分,以获取数量较少的能够覆盖像素区域的矩形.实验证明,改进的算法在划分矩形时能极大地减少矩形数目,提高压缩效率.