Fabric Defects Detection based on SSD

Abstract

In this paper, Fabric defect detection is a challenging task because of the complex texture. Deep learning technology provide a promising solution. As a kind of deep learning object detection model. Single Shot Multibox Detector(SSD)achieves good detection performance. However, the original SSD model may fail to detect the small objects. In this paper, we proposed a novel SSD model for fabric defect detection. Experimental results showed that the improved SSD model can accurately detect the defect region.