On-line Detection Of Paper Defect Based On Image Processing Technology

Along with the innovation of papermaking technology, the speed of papermachine is faster, paper banner is wider. Artificial paper defect detecting methodwidely used currently cannot satisfy the rapid development situation of modernpaper industry. Therefore, an intelligent paper defect online detecting system isurgently needed for long-term development of paper industry. In this paper, anonline detection system for paper defect with image processing technology wasdesigned and studied.According to the system anticipative index requirements and papercharacteristics, performance parameters of equipments used in light sourcesystem, image acquisition system, image processing system and auxiliary systemwere analyzed in detail. Hardware system of the online paper defect detectionsystem was built with suitable components.Filter algorithms, image segmentation algorithms and feature extractionalgorithms were analyzed. And simulation experiments were carried out. Asonline paper defect detecting system has the characteristics of high real-time andlarge amount of information, median smoothing filter algorithm of3by3template was used to filter the noise of the paper image. Paper image wassegmented using the adaptive threshold segmentation method and the iterationthreshold segmentation method. Paper defect contour information was extractedwith boundary tracking. Common types of paper defects can be recognized afterthe geometry and gray characteristics of paper defect were extracted andanalyzed.Software system of online paper defect detecting system was completed onMicrosoft Visual C++6.0platform. Functions and development process of everymodule of the software system were analyzed. And detection algorithms wereintegrated into the software system.The online paper defect detecting system was verified while light system and image acquisition system under stable performance situation. Theexperimental results show that the detecting system can achieve the anticipativetechnical requirements. Common paper defects can be detected and recognizedsuccessfully.