基于Hough变换和颜色特征的预瞄准技术研究

Study on Precollimation Based on Hough Transforms and Color Character

为实现激光通信终端地面验证实验中终端的自动对准,提出了一种基于图像识别的预瞄准方法。该方法首先基于颜色特征提取出图像中的指示光斑与标尺光斑,根据标尺光斑的尺寸,建立图像与实物的比例关系,并预估出目标圆的半径像素值。然后,基于Hough变换原理识别出图像中的通信终端通光中心位置,并与指示光斑进行比较计算出补偿角度。将补偿角传给运动控制卡,控制指示光瞄准目标终端。最后,启动玫瑰曲线扫描,捕获信标光后完成预瞄准过程。实验结果表明:该方法能准确识别目标,计算时间为0.94 s,并具有良好的抗干扰能力。基本满足通信终端自动识别、快速对准的设计要求。

In order to realize auto-alignment of terminal in ground laser communication simulation experiment, a precollimation method based on image recognition is proposed. First, indication cursor are extracted and reference cursor according to the color character of image. On the basis of the size of scale facula, the method establishes proportion relation between image and real object is built, and then the radius of target circle is estimated. Then based on Hough transform, the algorithm finds the centre of communication terminal caliber in the picture and calculates compensation angle between the centre and indication cursor, by sending the compensation angle to motion control card, it would make laser directly collimate to the terminal. At last, starting rose scanning and the process is terminated until capturing beacon light. Experimental results indicate that this method can detect the target exactly, and the calculation time is 0.94 s. It has perfect anti-jamming character. The method satisfies the requirements of auto-recognition and rapid collimation between two communication terminal.