粗糙集和规定化直方图图像增强算法研究

图像增强能够有选择地突出图像中的有用信息并抑制无用信息，以改善图像视觉效果进而提高图像清晰度和使用价值。由于图像信息本身具有复杂性和相关性，导致在图像增强中出现不确定性和模糊性。提出粗糙集和规定化直方图的图像增强算法，采用粗糙集表示图像，用规定化直方图和最大类间方差法求出划分图像的灰度阈值和噪声阈值；将图像划分为背景子图、前景子图和噪声子图，分别对各个子图进行增强处理，然后合并子图以增强图像。实验分析表明，粗糙集、规定化直方图和最大类间方差法相结合的图像增强处理对灰度图像有效，具有一定的实用性。

Image Enhancement Algorithm for Rough Set and Specified Histogram

Image enhancement can selectively highlight useful information in an image and suppress unwanted information to improve image visuals and thereby improve image clarity and usefulness. The image information itself is complex and correlated， leading to uncertainty and ambiguity in image enhancement. An image enhancement algorithm for rough sets and prescribed histograms is proposed. The image enhancement algorithm uses a rough set to represent the image and then obtains the grayscale threshold and the noise threshold of the divided image by using a prescribed histogram and a maximum inter-class variance method， and divides the image into a background subgraph， a foreground subgraph， and a noise subgraph. The processing of each subgraph is enhanced and the subgraphs is merged to enhance the image. Through experimental analysis， the rough set， the specified histogram and the maximum inter-class variance method are combined to enhance the image， which shows that the enhanced algorithm is practical and effective for grayscale images.