摘要
图像识别技术在移动终端设备中的应用已经是普遍趋势。文章针对目前的应用趋势,基于OpenCV框架,设计并实现了一个具有预分类识别机制的综合性图像识别系统。系统在具体识别前,对图像分割处理,使图像转化为大量离散子图像,进而对子图的HSV、直方图、熵值、轮廓等特征值的匹配分析,达到对图像中多ROI对象快速分类识别的目的。根据预分类结果,调度二维码、车牌等识别库,实现图中多对象的识别。相比无此机制的系统,其识别效果可明显得到改善。文章设计的系统平台适用于图形识别领域人员进行图形识别学习研究,对复杂对象识别场景也有较高的参考应用价值。
The application of image recognition technology in mobile terminal equipment has been a general trend.In view of the current application trend,this paper designs and implements a comprehensive image recognition system with pre-classification and recognition mechanism based on OpenCV framework.Before the specific identification,the system performs image segmentation processing to convert the image into a large number of discrete sub-images,and then matches and analyzes the feature values such as HSV,histogram,entropy and contour and other feature values of the sub-images,so as to achieve the purpose of rapid classification and recognition of multiple ROI objects in the image.According to the results of pre-classification,the recognition libraries such as QR code and license plate are scheduled to realize the recognition of multiple objects in the figure.Compared with the system without this mechanism,its recognition effect can be significantly improved.The system platform designed in this paper is suitable for people in the field of graphic recognition to study graphic recognition,and has high reference application value for complex object recognition scenes.
作者
邱敏
朱启文
Qiu Min;Zhu Qiwen(Department of Computer Management,Nanjing Technical Vocational College,Nanjing 210019,China;School of Electronic Information,Nanjing Vocational College of Information Technology,Nanjing 210023,China)
出处
《信息化研究》
2021年第6期48-51,64,共5页
INFORMATIZATION RESEARCH
关键词
预分类
图像识别
图像分割
pre-classification
graphic recognition
image segmentation
