The Research Of Image Detection System Of Fiber Tape Laying Based On Machine Vision

Due to fiber composite material's many advantages, such as low density, high strength, and excellent resistance to corrosion, they are now widely used in military industry ,aerospace, automotive industry, etc. Automated tape laying is a mature process and is currently being used in commercial applications. It is essential to detect the angle of the fiber laid down and the spacing between the fibers. Experiments prove that the angle of the fiber laid down could affect the mechanical property of the composite material. To meet the specified design goals, the adjacent bands on the lay-up surface must be close together and there must be no gaps or overlaps between the bands.The technology, which used to construct the detection system of the bands' angle and gap based on machine vision, are present in this paper. The detection system based on machine vision always has many advantages such as high speediness, high precision, contactless measure, etc. In some cases, it can replace the inspection of the people. The detection system constitutes of the USB cameral as the image sensor and the PC which is designed as the system processor. So the cost is low and the cost performance is high.The system used the VFW technology to capture the video sequence of the laying bands through the USB cameral. The image process is the emphasis of the detection system. To enhance the quality of the image, the image is processed by the medium filter and histogram equalization algorithm. To improve the real time performance of the system, a fast algorithm of medium filtering is adopted. The color image segmentation based on pyramid link approach greatly improve the band's recognition rate. The detection of edge of band may use many kinds of operators, then methods were introduced to raise the edge's precision to sub-pixel. Here, three ways of measuring the angle of the bands in the image processed by thinning algorithm are raised, fitting method, rotating and projection way and fast dynamic Hough transform way. And two ways of detecting the gap between adjacent bands are introduced, fitting method, and projection center way. Their speed and precisions are tested by experiments respectively. To get the relation between the pixel distance in image and actual range, the program of camera calibration is introduced and the error caused by camera's distortion is estimated.At last, on basis of MFC and OpenCV, the software for the tape lying detection is realized. The study of the technique which used to construct the image detection system of tape laying ,is of much practical value.