Abstract: Because of the characteristics of the container itself and the error of X-ray machine, the container X-ray images have lots of strong noise. In order to improve the region of interest and inhibit the region of uninterested in, an improved method based on contrast limited adaptive histogram equalization was introduced in the paper. Firstly，the X-ray images were enhanced by using a method of contrast limited adaptive histogram equalization, and then a method of median filter was applied to the processed image in order to reduce the noise, in order to get the low-frequency information of the X-ray images. Secondly, the relatively high-frequency information of the original image was got by Laplacian filter. Finally, the weighted sum of the previous two processed images was the goal image. Experimental results show that the method can largely reduce the strong noise, and preserve the image information interested in better, and help the study of container X-ray image.

Key words: image enhancement, histogram equalization, local histogram equalization, contrast limited adaptive histogram equalization

摘要： 由于集装箱自身的构造特点以及X光扫描机的信息采集误差，使得集装箱X光图像有很强的竖纹噪声。为提高X光图像中感兴趣区域的清晰度并抑制不感兴趣信息，提出了一种以有限对比适应性直方图均等化为基础的改进方法：对原始图像进行有限对比适应的直方图均等化处理，进行图像增强，并对处理后的图像进行中值滤波，以减少箱体的竖纹噪声，得到图像的低频信息，然后，对原图像进行拉普拉斯滤波处理，得到图像的相对高频信息。最后，将处理后的两幅图像加权相加，作为增强后的目标图像。实验证明，该方法能有效削减竖纹噪声，很好地保存图像的目标信息，便于对集装箱X光图像进一步研究。

关键词: 图像增强, 直方图均衡化, 局部直方图均衡化, 有限对比适应性直方图均等化