摘要
水面无人艇作业过程中,运动目标是其重要的信息参考来源。为了提高水面运动目标检测效率,研究了多种运动目标检测方法。在此基础上,研究了一种有效的水面无人艇动态目标检测方法。首先,以改进的高斯差分对图像进行处理。其次,结合三帧差分法的原理对图像进行运动目标检测。最后,利用数学形态学原理对运动目标进行更好的提取。实验结果表明,该算法可以较好地进行运动目标检测,且算法复杂度低,易实现。
During the unmanned surface vehicle homework process, moving object is the important source of information reference. In order to improve the efficiency of the water moving target detection, a variety of moving target detection method is studied. On this basis, a kind of effective method to detect dynamic target, unmanned craft on the surface of water is studied. First of all, the improved Gaussian difference of image processing is carried out. Secondly, moving object is detected by combining with the principle of three frame difference method. Finally, the motion target is well extracted by using mathematical morphology theory. Results show that the algorithm can be better for moving target detection, with low algorithm complexity and easy implementation.
出处
《沈阳理工大学学报》
CAS
2016年第5期33-38,共6页
Journal of Shenyang Ligong University
关键词
三帧差分
高斯差分
形态学
three frame difference
gaussian difference
morphology