无限靶位重叠图像有用信息提取方法

Infinite Overlapping Target Image Method to Extract Useful Information

研究一种无限靶位重叠图像的引入反常位移算子去伪处理方法。从无限靶位重叠图像中提取有效信息是非常困难的。传统的无限靶位重叠图像处理方法采用图像单帧处理的方法,对恶劣重叠情况下的处理结果较差。提出一种无限靶位重叠图像的引入反常位移算子去伪处理方法,采用引入反常位移算子去伪处理的方法,对无限靶位重叠图像在多个维度进行相关信息融合和识别实现,提取图像的多维综合信息,通过引入反常位移算子去伪处理,实现无限靶位重叠图像信息的去伪处理。采用一个重叠非常严重的靶位图像进行实际的去伪测试实验。结果显示,采用引入反常位移算子去伪处理方法,无限靶位重叠图像中的有用信息被很好的提取出来,去伪效果明显,方法将在运动循环靶射击训练和瞄准仪开发等方面具有很好的应用价值。

The camouflage remove of infinite overlapping target image with introduction of abnormal displacement operator was researched. It was very hard to extracting useful information from the overlapping images. In conventional unlimited target overlay image processing method, a single frame of the image processing method was used, and the worst case of overlapping processing result was poor. The camouflage remove of overlapping image with introduction of abnormal displacement operator was proposed, the anomalous shift operator dummy processing method was introduced, the superimposed images of the infinite target multiple dimensions and identification of relevant information was extracted to achieve fusion, multi-dimensional integrated information, and through the introduction of pseudo-anomalous shift operator, the unlimited image information for target overlapping pseudo handle was achieved. A serious overlapping target image was used to do the actual test experiment, the result shows that with the introduction of abnormal displacement operator, the unlimited target overlapping images is extracted well, the effect is obvious, and it has good application value in gunnery training and aiming rule sight device research.