基于信息熵的X射线弹药图像自适应局部模糊增强

An Adaptive Local Fuzzy Enhancement Approach Based on Information Entropy for X-Ray Ammunition Image

针对X射线弹药图像对比度较低、缺陷边缘模糊的特点,提出了一种基于信息熵的自适应局部模糊增强算法。该方法先对弹药射线图像进行动态归一化处理,并运用梯度算子获取感兴趣区域,OTSU算子自动选取最佳渡越点gc,从而得到模糊增强算子的渡越点cμ,再使用改进的隶属函数对cμ两侧的像素灰度进行模糊域的非线性处理,从而得到增强后的弹药图。试验结果表明,该算法能明显提高弹药图像的对比度,突出疵病并降低背景噪声。

For the problems of the low contrast and fuzzy flaw edge in X-ray ammunition image,an adaptive local fuzzy enhancement approach based information entropy was presented.It first dynamically normalized the ammunition image,obtained the regions of interest by gradient operator,and automatically selected the best crossover point gc with OTSU algorithm in order to get the transition μc of fuzzy enhancement operator.It then used the improved membership function to nonlinearly fuzzy process the gray value on the two sides μc to obtain the enhanced ammunition image.Experimental results demonstrated that the proposed approach could efficiently improve the contrast of the ammunition image and enhanced the flaw information.