【摘要】 为实现对棉花异性纤维自动视觉检测系统采集的彩色图像的精确分割,提出了基于视觉显著图的异性纤维彩色图像分割方法。通过计算颜色显著图,实现彩色异性纤维目标的识别;通过计算亮度显著图,实现灰色异性纤维目标的识别;将彩色和灰色目标进行融合,得到全部异性纤维目标。实验结果表明,该方法可以准确分割出异性纤维彩色图像中含有的各种异性纤维目标。通过比较发现,该方法在分割精度上明显优于其它方法,可以实现对异性纤维彩色图像的精确分割。更多还原

【Abstract】 It is difficult to separate foreign fiber objects from background in a live image captured by an automated visual inspection system for foreign fiber detection duo to the inhomogeneous brightness of background and various types of foreign fibers in different colors and shapes.A saliency map based color image segmentation method which aims at foreign fiber detection is presented.The RGB color image captured in real-time is firstly separated to R,G and B color channels.Then the red,green and blue color features are calculated respectively from the corresponding R,G or B channel.Afterwards,three saliency maps are obtained from these three color features and then fused together.The fused saliency map is segmented to get the color foreign fiber objects.Those foreign fiber objects in dark black or bright white are also segmented out using a threshold method from the brightness saliency map.At last,all obtained foreign fiber objects are fused together to generate the final objects.The results indicate that the proposed method can segmented out color foreign fiber objects as well as gray foreign fiber objects in dark black or bright white.Through result comparison,it indicates that the proposed approach is significantly better than other methods in segmentation precision,and can realize the segmentation of foreign fiber color images.更多还原