天空
红外线的
遥感
计算机科学
环境科学
天文
地质学
物理
作者
Anbo Jiao,Liang Xue,Xinke Lian,Tianru Dai,Penghao Yan
摘要
Aiming at the problems of large amount of calculation and insufficient detection accuracy in the existing infrared ship detection methods with sea and sky background,a method for fast detection of infrared ship targets using projection method is proposed. Firstly,the median filter is used to smooth the image to improve the signal-to-noise ratio,and then the infrared image is converted from gray image to binary image to obtain edge information. Secondly,the edge points are sampled in the edge image by region, and a straight line is fitted to detect the sea antenna in the image. Finally,the sea antenna in the edge image is eliminated,so that the ship target can be better separated. When the position and size of the ship target are detected,the projection method is used for dimension reduction calculation to improve the speed. The edge points of each line of the image are accumulated in the horizontal direction to screen out the range of potential targets. Then the edge points of each column of the image are accumulated in the vertical direction to narrow the range of potential targets. Because multiple ship targets may exist in the horizontal or vertical direction of occlusion,the second projection can be carried out in the horizontal and vertical directions to eliminate the occlusion and further narrow the range of potential targets. The shape of the target is represented by the ratio of the length and width of the area where the potential target is located, the size of the target is approximated by the area of the potential target area, and the location of the target is approximated by the center of the potential target area. The proposed method has small amount of calculation, and can improve the real-time performance of ship target detection. Experimental results show that the proposed method is suitable for ship target detection in infrared imagery with sea and sky background, and has high detection accuracy and fast detection speed.
科研通智能强力驱动
Strongly Powered by AbleSci AI