全景图像海天线提取及舰船目标自动检测

Sea-sky Line Extraction and Ship Target Automatic Detection of Panoramic Image

为实现海洋环境下全景摄像机自动目标检测,提出了一种全景图像海天线提取算法及海天线上舰船目标检测方法;首先,分析了全景图像中海天线的成像特点,使用基于分区的自适应阈值Canny边缘检测算法进行边缘检测,并对海天线边缘进行双阈值梯度方向过滤;然后,对图像边缘进行细化,提出了基于最长曲线法的海天线边缘提取算法并进行椭圆拟合得到海天线边缘成像椭圆方程;最后,介绍了根据海天线椭圆方程对海天线上舰船目标进行检测的方法;使用3种不同海洋环境下拍摄的图像进行了实际测试,实验结果表明:该方法可有效地检测出复杂海天背景下的海天线及海天线上舰船目标,海天线提取成功率可达95%以上,对径向成像高度超过10个像素的目标识别成功率可达90%。

In order to realize automatic ship target detection with panoramic camera in the marine environment, a sea--sky line extraction algorithm for panoramic image and a ship target detection method around sea--sky line was proposed. Firstly, the characteristics of sea--sky line in panoramic image are analyzed. An improved Canny edge detection method based on partition adaptive threshold was used, and double threshold gradient direction filtering was carried out. Then after edge thining, the sea--sky line edge extraction algorithm was proposed based on the longest curve method and the ellipse fitting was performed to get the sea--sky line elliptic equation. Finally, the ship target on sea--sky line detection method was introduced based on the sea--sky line elliptic equation. The aetual test was carried out using real panoramic images obtained in three different marine environment. The experimental results prove that this method can effectively detect the sea- sky line and the ship target under complex sea--sky background. The success rate of sea--sky line extraction can be up to 95% and the detection success rate can reach 90% when the ship target radial imaging height is more than 10 pixels.