设施农业喷雾机器人的组合视觉导航方法

Visual navigation method of agro-spraying robot based on navigation line and QR code combination

温室环境下,针对喷雾机器人在直线路径行驶时会存在偏差,在转弯路径处又需迅速获得导航信息的情况,设计一种导航线和QR码组合的视觉导航算法.在直线行驶路径贴上红色导航线,对采集到的图像处理后获得偏差、偏角进行方向修正,实现直线行驶精准导航.但导航线导航在转弯行驶时存在转弯效率低、偏差大等问题,因此设计QR码导航方法,通过识别贴在转弯处的存储有导航和决策信息的QR码,完成转弯处的迅速精准导航的要求.采集各种工况下导航线和QR码图像各120幅进行算法验证,导航线的平均纠正率为93.33%,QR码平均识别率为92.50%.导航试验结果表明:行驶速度在2.0 m·s^(-1)时,导航线和QR码的最大导航偏差分别为2.0和3.5 cm,满足喷雾机器人导航的实时性和准确性要求.

In the greenhouse environment, a navigation algorithm with navigation line and two-dimensional code was designed to solve the deviation in the straight path of spraying robot and quickly obtain the navigation information at the turning path. The red guideline was pasted on the straight-line driving path, and the collected images were processed to calculate deviation and declination and converted into right-and-left-wheel differential for correction to achieve accurate linear navigation. To solve the problems of navigation-line navigation with low turning efficiency and large deviation during cornering, a QR code navigation method was designed to realize prompt and accurate navigation by recognizing the QR code at turning point with navigation and decision information. 120 navigation line images and 120 QR code images under various conditions were collected to verify the algorithm, and the average correction rate of navigation line and the average recognition rate of QR codes were 93.33% and 92.50%,respectively. The results of navigation experiments show that the maximum navigation deviation of navigation line is 2.0 cm, and the maximum navigation deviation of the QR code is 3.5 cm at the traveling speed of 2.0 m·s^-1 , which can meet the real-time and accuracy requirements of the navigation of spraying robots.