基于图像分析的后视镜失真率自动检测方法

传统的机动车后视镜失真率检测方法采用模拟相机拍照、照片冲洗、人工测量等步骤，效率低下且人为因素对测量结果影响较大。本文使用数字相机成像，针对数字图像，设计并实现了基于图像分析的失真率自动检测算法，通过圆心检测、图像分割、直线与圆交点检测等几个步骤，完成了自动检测功能。该方法高效、准确，达到了实用水平。

Automatic Distortion Detection for Automobile Rearview Mirror Based on Image Analysis

In the distortion detection for automobile rearview mirrors, the traditional method is photographing with an optical camera, developing photographs and measuring artificially. Due to the influence of artificialness, it is inefficient and inexact. This paper introduces a new automatic detection method based on the computer analysis of the images from digital camera. By the steps of circle centre detection, image segmentation and cross detection, the method succeeds in detecting distortion automatically, efficiently and exactly. Experiments show it is effective in actual applications.