计算机视觉快速开发框架的研究与应用

Computer Vision Research and Application of Rapid Development Framework

在计算机视觉分析应用项目中,常采用多种算法进行实现。在特定的环境下,如何找出其最优路径,往往需要通过大量的对比实验。而应用场景每变化一次,需要进行多次重复实验,不但延长了开发时间,同时也增加了开发成本。计算机视觉快速开发框架从实际应用工程的角度出发,将视觉技术的核心算法从底层研究中剥离,通过扁平化、模块化、一体化的处理、任意组合以及对数据流的合理配置,使应用项目中的开发人员各司其职,分工明确,提高了开发结果的复用性,缩短了开发时间,提高了开发效率。

In computer visual analysis application projects,often using a variety of algorithms to achieve.In certain circumstances,how to find the optimal path,often required by a large number of comparative experiments.Once every change in the applica tion scenario,the need for repeated experiments,not only to extend the development time and development costs also increased.Computer vision rapid development framework from the point of view of the practical application of engineering,visual technol ogy core algorithm is stripped from the underlying research through the flat,modular,integrated processing,in any combination,as well as the rational allocation of the data stream,so that the application project the developers to carry out their duties,a clear division of labor,improve the reusability of the results of the development,shorten the development time and improve develop ment efficiency.