摘要
本设计的目的是为了在现有的颜色特征识别系统中添加“自动提取特征–识别”的特性,在ROS (Ubuntu)环境下构造一个具有独立图像数据收发接口和指定框架的特征文件读写接口的特征提取归档模块。通过集成Opencv、XmlLib2的基本功能函数,自制图像收发接口类、特征文件读写类,并为特征提取编写了以RGB颜色模型为基础的提取算法使该模块能够适配ROS环境。在经过数次静态测试后,本设计的提取归档模块基本达到了对指定区域的等比例抽样提取、按累计数量筛选特征、将特征归档于指定目录下的设计目标。
The purpose of this design is to add the feature of"automatic feature extraction-recognition"into the existing color feature recognition system,and construct a feature extraction archiving module with an independent image data transceiver interface and a feature file read-write interface of the specified framework in the ROS(Ubuntu)environment.By integrating the basic functions of Opencv and XmlLib2,we made our own image transceiver interface class and feature file read-write class,and wrote an extraction algorithm based on RGB color model for feature extraction,so that this module can adapt to ROS environment.After several static tests,the extraction and archiving module of this design basically achieves the design goal of sampling and extracting at the same proportion of the designated area,screening features according to the cumulative number,and archiving features in the designated directory.
出处
《图像与信号处理》
2019年第3期155-162,共8页
Journal of Image and Signal Processing
基金
天津市高等学校科技发展基金资助项目(项目编号:JWK1614)。
