不变矩在图像特征提取及目标识别中的应用

Application of Moment Invariant Theory in Digital Image Character Recognition

模式识别是人工智能的一个重要分支,是当代高科技研究和应用的重要领域之一。围绕数字图像识别这一中心课题,研究了目标图像的预处理、不变性特征的提取方法。最后对目标图像的识别进行了计算机仿真实验,结果证明以上方法是可行的。

Pattern recognition is an important branch of artificial intelligence, and it is one of the important realms of high technique research at present. Making use of its dominance in information handling to solve the problem of image identification is the purpose of this paper. The approach to the pretreatment and invariant feature extraction of the target image is investigated. Finally, the simulation experiments on the target image identification are carried out. Experimental results indicate that the method is feasible.