尺度旋转相关滤波视觉伺服研究

Research of Scale and Rotation Based Correlation Filter Visual Servo

为了满足工业现场日益增多的应用需求,提高视觉伺服目标跟踪算法应用性,研究并提出了一种尺度旋转相关滤波视觉伺服实现方法。通过对MOSSE、ASET等先进算法进行改进,利用仿射原理,嵌入尺度旋转独立滤波器,构建适合工业现场的尺度相关滤波视觉伺服实现方法及实验平台,该实验平台由视觉伺服和二维运动控制平台组成,视觉伺服的输入为实时图像,输出为机器人动作;二维运动控制平台准确驱动被跟踪的目标,使跟踪结果可以量化。实验证明:该视觉伺服实现方法在改进鲁棒性和实时性的同时,可以满足工业现场应用需求。

In order to satisfy the increasing demand of industrial field and to improve the applicability of visual servo target tracking algorithm,a scale rotation correlation filter based visual servo implementation method is researched and designed.By improving the MOSSE,ASET and other advanced algorithm,affine principle was embedded into scale rotation independent filter to construct a scale related filter visual servo implementation method and experimental platform which is suitable for industrial field.The experimental platform was composed of visual servo and two-dimensional(2D)motion control platform.The input of visual servo was real-time image,and output was robot action.The 2D motion control platform accurately driven the target being tracked,so that the tracking result could be quantified.The experimental results show that this method improves the robustness and real-time performance,and can meet the needs of industrial field application.