基于字符识别的收割机机器视觉研究

Research on Harvester Machine Vision Based on Character Recognition

随着农业现代化技术的不断发展,无人驾驶技术开始被应用到农机的设计上,为了实现农机的无人驾驶,首先需要使农机具备智能化视觉系统。为了提高农机视觉系统的导航线识别能力,将基于翻译工具的字符识别技术引入到了系统的设计上,结合图像处理技术,通过对导航线的初步定位、二次定位和图像的滤波、灰度化及增强处理,可以准确识别出作业区域的导航位置线,还可以得到实时的位置坐标。测试结果表明:收割机在自主行走时可以沿着导航线位置行走,并通过坐标点来实时的修正行走方向,达到了精准定位导航的目的。

With the continuous development of agricultural modernization technology,unmanned driving technology has been applied to the design of agricultural machinery.In order to realize the unmanned driving of agricultural machinery,it is necessary to make agricultural machinery have intelligent vision system.In order to improve the recognition ability of navigation line in agricultural machinery vision system,it introduced the character recognition technology based on translation tool into the system design.Combining with image processing technology,the navigation line can be accurately identified by preliminary positioning,secondary positioning and image filtering,graying and enhancement processing,and real-time positioning.The harvester can walk along the navigation line position when walking independently,and correct the walking direction in real time through coordinate points,so as to achieve the purpose of precise positioning and navigation.