Allocation And Recognition Of The Anchor Point In PCB Board

The technique of image interpolation has important research value. In this paper, BP neural network, wavelet transform, linear interpolation are applied to image interpolation in order to improving image resolution. The high resolution image acquired has better visual quality subjectively and higher value of PSNR objectively, besides, it retains details in the image better. The edge blurring and ringing artifacts have been obviously reduced. According to image interpolation, a novel method of recognizing the anchor point is proposed. Whether the anchor point image is round or not, all depend on arearatio and circularity. It also realizes the allocation of the circle anchor point correctly. Due to noise and uneven illumination, we must divide the original image of the anchor point into pieces, segment each subgraph using its own threshold value, and the binary image of the original image is obtained. Moreover, the binary image is curve fitted. Therefore, we can fix on the center of the circle anchor point. The results illustrate, the method about the allocation and recognition is effective.