一种基于模板匹配的远场畸变校正方法

在激光照射性能监测系统跟踪移动目标采集远场图像过程中,由于受主光轴入射角度、相对位移等因素影响,导致所摄图像会产生一定程度的畸变。为了校正图像畸变,提出一种基于模板匹配的远场畸变校正方法。首先制作四个靶板角点模板,利用模板匹配算法跟踪识别畸变图像的靶板角点;依据相机成像模型计算出理想靶板角点的像素坐标;根据四对靶板角点,利用透视投影模型求解畸变校正系数,从而实现对图像的畸变校正。实验表明,该方法可以对远场畸变图像进行有效校正,具有实验条件要求低、校正精确、处理速度快等优点,为激光照射性能监测系统提供了准确的图像信息,在实际工程中具有重要意义。

A Far-field Distortion Correction Method Based on Template Matching

In the process of tracking the moving targets and capturing the far-field image with the laser radiation performance monitoring system,due to the incident angle of main optical axis and the relative displacement and other factors,some degree of distortion will be produced in the captured image.In order to correct the image distortion,a far-field distortion correction method based on template matching is proposed.Firstly,corners templates of four targets are made;corners of target in the distorted image are tracked and recognized by using template matching method;pixel coordinates of the ideal target corners are calculated according to camera imaging model.According to the four corners of the target,the distortion correction coefficient is solved by using the perspective projection model in order to achieve the correction of the distortion image.Experiments show that the far-field distortion image can be effectively corrected by using the method,and it has advantages such as requiring low experimental conditions,precise calibration,high processing speed,etc..The accurate image information can be provided for laser radiation performance monitoring system and has an important significance in the actual project.