X射线异物检测的图像屏蔽算法

The Application of Image Shielding Method in the X-ray Foreign Material Detection

X射线异物检测系统检测包装食品过程中,由于产品某些包装区域的材质以及厚度特征,此区域对X射线的吸收率与异物对X射线的吸收率十分相似,导致系统在异物识别时将这些区域误判为异物,造成误检。针对此类问题,结合X射线异物检测原理和图像分割理论,提出了4种图像屏蔽方法。Matlab仿真结果表明,针对不同检测产品采用相应的屏蔽算法,均可有效屏蔽该检测产品中被误判为异物的区域,大大提高了X射线异物检测系统的灵敏度,具有较好的商业使用价值。

In the X-ray foreign material detection process, it is often misjudged some parts of the packaged products as foreign bodies, because the X-ray absorption in these parts are similar to that in the foreign bodies. However, the problem can be well solved by using image mask methods. To this problem, this paper put forward four kinds of image mask methods by combining the basic theories of image segmentation and the detection principle of X-ray foreign material inspection system. Matlab simulation results show that those error detection parts can be masked effectively by using corresponding mask method. And to some extent, it is proved that the proposed mask methods improve the sensitivity of the X-ray foreign material detection system and have good commercial use value.