直方图均衡技术综述①-计算机系统应用.PDFVIP

2014 年 第 23 卷 第 3 期 计 算 机 系 统 应 用 直方图均衡技术综述① 李乐鹏, 孙水发, 夏 冲, 陈 鹏, 董方敏 (三峡大学 智能视觉与图像信息研究所, 宜昌 443002) 摘 要: 直方图均衡是一种简单而有效的图像增强技术, 在医学图像处理、计算机视觉、遥感影像等领域都有重 要作用, 从均衡时选取的区域情况分为全局和局部直方图均衡方法; 从均衡时选取的像素值域可以分为传统直 方图均衡、双直方图均衡和递归分解的直方图均衡, 甚至多直方图均衡; 从均衡对象可以分为空域的直方图均衡 和结合频域滤波的直方图均衡; 从均衡图像的通道数可以分为灰度图像直方图均衡和彩色图像直方图均衡. 本 文对以上涉及的直方图均衡技术进行了较为系统的综述, 对衡量直方图均衡效果的评价指标进行了说明, 展望 了直方图均衡技术进一步研究方向. 关键词: 直方图均衡; 双直方图均衡; 局部直方图均衡; 彩色图像均衡 Survey of Histogram Equalization Technology LI Le-Peng, SUN Shui-Fa, XIA Chong, CHEN Peng, DONG Fang-Min (Institute of Intelligent Vision and Image Information, China Three Gorges University, Yichang 443002, China) Abstract: Histogram Equalization is a simple and effective technology in image enhancement, and it is widely used in medical image processing, computer vision and remote sensing image. When selected different equalization regions, histogram equalization can be divided into global and local histogram equalization method. While the equalizations choose different pixel area, it will be divided into traditional histogram equalization, bi-histogram equalization, recursive mean-separate histogram equalization, and even multi-histogram equalization. Histogram equalization also can be divided into histogram equalization based on space domain and histogram equalization combined with frequency domain in equalization objects. From the channel numbers of equalized image, it will be divided gray image equalization and color image equalization. This paper systematically summarizes the histogram equalization methods mentioned above, describes the evalua