摄像头阅卷系统图像预处理算法研究

本文讨论了摄像头自动阅卷系统中的图像预处理算法,对答题卡图像预处理过程中的图像灰度化、二值化、光学畸变校正和图像偏斜纠正等关键步骤进行了描述。在灰度化过程中采用彩色模型转换的方法获得灰度图,在二值化过程中采用由用户指定阈值的方法,在畸变变换中采用二元多项式法,在偏斜纠正中使用Hough变换检测倾斜角度。

Pre-processing Algorithm for Camera-based Information-card Review System

This paper discussed a series of pre-processing theory and algorithm for camera-based information-card review system. Converting RGB to grayscale image, image binarizing, optic aberrance revising and skew angle correction of image were described. In this paper, the way converting the deferent color model to get grayscale image was discussed. In the process of image binarizing, the threshold was decided by the user. A duality-multinomial method was presented in optic aberrance revising. Hough transform was applied to the skew angle correction.