基于Contourlet变换的火炮内膛疵病分割方法

对火炮内膛图像进行有效的疵病分割是内膛疵病识别的重要步骤。针对火炮身管导气孑L疵病的分割进行研究，提出基于Contourlet变换的火炮内膛疵病分割方法。通过对内膛图像Contourlet变换高低频系数的调整，在排除导气孔自身结构的影响下，提取出疵病的边缘，并结合形态学处理，实现导气孔疵病的分割。实验结果表明，该方法内膛疵病分割准确，抗噪能力强，适用于火炮内膛疵病分割。

Flaw Segmentation of Gun Bore Image Based on Contourlet Transform

The segmentation of gun bore flaw image is an important step in flaw recognition. In this paper, taking the segmentation of bore gas vent image as research object, a new flaw segmentation of gun bore based on Contourlet transform is proposed. By manipulating the high-low coefficients of image Contourlet transform, the flaw edge is detected while eliminating the influence of gas vent structure and the flaw segmentation is achieved through using the mathematical morpholo- gy. The experimental results show that the flaw segmentation using this algorithm has high accuracy and strong anti-noise ability that are fit for gun bore flaw segmentation.