Up to now, the aided image detection of containers with the X-ray inspection equipment is the main forms of regulation for entry and exit in customs in the process of customs. But the analysis and recognitions of container images is primarily dependent on the manual inspections. In order to improve the efficiency of supervision, the automatic analysis on container images is on in this paper and the goods in the containers are regular and fewer. Firstly, an improved method of based on contrast limited adaptive histogram equalization is introduced in my paper in order to enhance the image effectively, and the method can largely reduce the strong noise. Then extracting the objects with a method of mask and the threshold in gradient, and making a morphological processing after extracting the object, which can reduce the smaller noises. Finally, we can mark the illegal objects with the rectangular frames. And the whole processing can be helpful to reduce the manual interventions and to improve the supervisions. |