基于模糊主色调的彩色图像分割及道路检测

该文提出了一种新的基于模糊主色调的彩色图像分割方法。其主要内容是利用各类模糊主色调在各属性类中所占比例的不同构造一动态色彩查找表，并由像元的色彩值求得所对应的模糊主色调值，再由该值查色彩表确定其所属的类别。该方法合理地避免了有监督模式分类法所固有的缺点，以及无监督模式分类法对同一类中色彩的一致性要求。试验表明该方法速度快，效率高，并且对噪声不敏感。

An Approach to Color Image Segmentation Based on Fuzzy Dominant Color and Road Detection

A new approach to color image segmentation based on the fuzzy dominant color is proposed.According to the proportions of all the fuzzy dominant colors in each area of image,a dynamic color look up table is constructed.The fuzzy dominant color is computed for every pix of current image.And the classification labels are marked through the look up table.The approach avoids the intrinsic drawbacks of supervised classification methods and the homogeneous requirements of each color in unsupervised methods.Experimental results show that the new approach is effective and insensitive to noise.