Research on Ship Target Recognition based on Infrared Image Method

Abstract

Infrared thermal imaging technology has been widely applied in the field of target detection at present, which lays the foundation of research on situational awareness for the Maritime Autonomous Surface Ship. Compared with the traditional image recognition technology with defects such as low recognition accuracy, strong light dependence and poor anti-interference ability, infrared thermal imaging technology is not only adaptive to different light intensity environments, but also has the advantages of high concealment, strong detection capability, long detection distance and high detection sensitivity. In this paper, it proposes an improved method of Canny segmentation algorithm based on maximum inter-class variance method on the basis of infrared imaging, where the image is preprocessed by wavelet transform, it is achieved the moving target detection, to restrain noise interference. And then, the edge blurring is effectively processed by using the mask image of Canny edge detection as inputs of pattern recognition. The optimal threshold is determined by the improved Otus algorithm, which can stabilize the optical flow field of the background, to realize effective segmentation of ship images to obtain the high definition moving target. Finally, the availability of the algorithm has been verified by taking the tourist ferry under the Yangtze River as the object. The results showed that the improved algorithm can stabilize the optical flow field of the background, and the application effect is improved, which can provide technical support for the research and application of intelligent ship target perception.