基于小波包与大津算法结合的花生种子检测方法

For the edge feature of peanut seeds is irregular, and traditional detection method restricted by structural element. So an edge detection method has been proposed, which based on the wavelet packet and otsu. First of all, the image of peanuts decomposed by using wavelet packet. Then selected the optimal wavelet packet of reconstruction. Using otsu detected the edge of low-frequency part. Then two part edge of image fused together. Getting the edge of image with high resolution. The result of experiment showed that this method could accurately and clearly extract the edge information of peanut seeds. Compared with the traditional edge detecting method. This method detect peanut seeds successful, which can decrease the noise and enhance the edge. Then the precision of detection has been improved.