摘要
针对基于几何变换映射的色盲矫正方法进行色盲矫正时出现的映射计算量过高的问题,一种新型的几何变换映射色盲矫正方法被提出,这种矫正方法是基于图像分割原理的。首先研究了色盲图像的仿真方法,然后结合K-means和系统聚类算法对原始图像进行分割并计算各个区域在色盲图像下LAB颜色空间中的欧氏距离作为颜色相似性的度量,确定红绿色盲难以分辨的颜色区域。在这个区域内进行颜色映射区域的划分,这个划分主要包括根据领域内的颜色比例对空间内的几何平面进行相应变换,通过此类变换生成有利于色盲患者分辨的图案等步骤。
Aiming at the problem that the color blindness correction method is too high when the color blind correction method based on geometric transformation mapping is used, a color correction method based on image segmentation is proposed. Firstly, the simulation method of color blind image is studied, and then the original image is segmented by K-means and system clustering algorithm. The Euclidean distance in LAB color space of each region is determined as the measure of color similarity. Distinguish the color area. Then, the images in the color region are transformed according to the color ratio in the field, and then the different color mapping regions are divided into different color mapping regions. Finally, the color images are generated by color transformation. The surface of the experiment can improve the resolution of colorless patients with difficult to distinguish the color, and have a faster calculation speed.
出处
《电子测试》
2017年第8期43-45,共3页
Electronic Test
关键词
色盲
图像处理
几何变换映射
图像分割
Color blindness
Image Processing
Geometric transformation mapping
Image segmentation
