摄像头阅卷系统关键技术的分析与应用

在分析各类图像阅卷系统的基础上，阐述了摄像头阅卷系统中需要解决的关键技术问题有：对光照不均的图像处理，透视变形图像的校正处理。提出对图像进行二值化处理时采用局部平均阈值解决光照不均匀的问题；采用的基于矩形角点几何变换的畸变图像校正算法，分两步完成图像的倾斜校正和透视校正，并给出了变换公式。实践表明系统中采用的算法处理速度快、畸变图像校正效果好，图像识别率高，达到实用要求。

Key Techniques of Camera-Based Information-Card Scoring System and Its Application

In analyzing the various types of image-based scoring system, the key techniques of camera-based information-card scoring system are given: uneven illumination of the image processing, lens distortion correction image processing. This paper proposes binarization image processing with the local average threshold to solve the problem of uneven illumination. It uses rectangular corner-based geometric aberra- tion transform image correction algorithm, which include two steps to complete the correction skew dis- tortion and perspective distortion. It also gives the transformation formula. Practice shows that this method performs well, and meets the functional requirements.