基于交叉熵的图像过渡区提取与压缩研究

Research on image transition region extraction and compression based on cross entropy

改进过渡区提取算法,将最小交叉熵应用于红外图像过渡区提取中,提出基于交叉熵的图像过渡区算法,并将该过渡区分割法应用于红外图像感兴趣区域的自动提取,最后提出了一种基于JPEG2000框架红外图像感兴趣区压缩方案进行分类压缩。经图像实验充分验证了该方法的有效性、实时性,其具有重要的应用价值。

To improve the performances of transition region extraction algorithm, this paper proposed a novel transition region algorithm based on minimum cross entropy. Then applied it to detect ROI automatically. Finally, realized the ROI compression scheme for infrared image on the frame of JPEG2000. It is proves that the method of image compression based on region of interest automatic detected is reliable effective, and significant in applications.