基于超像素的印刷品色差在线检测方法

针对彩色印刷行业中,印刷品色差在线检测准确率低、速度慢的问题,提出一种新的检测方法.基于超像素的思想,采用简单线性迭代聚类(SLIC,simple linear iterative clustering)算法,对具有相似特征的相邻像素进行聚类,形成结构紧凑、近似均匀的像素块,每个像素块即为一个超像素.用超像素代替像素块内多个相似像素,分别提取标准图像和待检测图像的颜色特征.再用CIEDE2000色差公式进行色差计算.实验结果表明,该方法在保证检测结果准确率的基础上,能够有效地减少数据计算量,提高检测效率.

On-Line Detection Method for Printing Image Chromatic Aberration Based on Super-Pixel

In order to solve the problems of low accuracy and slow speed of on-line chromatic aberration detection in color printing industry,a new detection method is put forward.Based on the idea of super-pixel,the adjacent pixels with similar characteristics are gathered together by the algorithm of simple linear iterative clustering,each image is segmented into super-pixels which are some compact structure and approximate uniform pixel blocks.The multiple similar pixels in a pixel block are replaced by super-pixel,and the color features of the standard image and the image to be detected are extracted.Finally,the chromatic aberration is calculated by CIEDE2000 color difference formula.The experimental results prove that the proposed method can effectively reduce calculated amount and improve the detection efficiency on the premise of the certain detection accuracy.