Contour Primitive Recognition of Planar Thin Sheet Metal Based on Feature Point

To improve the efficiency of planar thin sheet metal production and guarantee the quality of the sheet metal products which has a plurality of apertures, an automatic measurement system based on machine vision is applied to measure the sheet metal products' dimensions. For the measurement system, contour primitive recognition is one of the most important technologies. In this paper, a recognition algorithm for contour primitives of planar thin sheet metal based on feature point is proposed. The algorithm can be used to detect corner points of the contour and recognize the contour primitives. The procedures are as follows. First, the image of thin sheet metal is captured and the closed contours are detected. And then preliminary identification of the feature points of the contours is performed by using K-cosine algorithm. Consequently, the candidate feature points are modified according to DXF file and the accurate feature points are obtained. Finally, the contour primitives are recognized. Experiments are carried out to verify the validity of the proposed method.