Tea Ridge Identification and Navigation Method for Tea-plucking Machine Based on Machine Vision

Abstract:

In order to improve tea quality and efficiency of plucking by using the teaplucking machine, this paper proposed a kind of identification method of the tea ridge and navigation method of the teaplucking machine. Firstly, the inside and outside parameters of camera offline were calibrated, which were put in the front of teaplucking machine for getting the video information about the tea ridge. Secondly, the Gauss filter method was used to reduce noise jamming, and after customizing the rectangular window, the local OTSU thresholds method was used to segment image, the image was obtained by the camera. Then the points which belonged to the edge lines of the tea ridge were found and noted based on space constraint, the space constraint can improve the speed and the accuracy to find the points. Least square linear fitting method was used to fit the left and right edge lines of the tea ridge. Then the center line of the ridge was calculated according to the calibration results, and the state of the teaplucking machine, including the center line of the ridge and the offset of the teaplucking machine, was showed on the screen of the driver’s seat in order to prompt the driver how to adjust the machine. Experimental results show that this method can identify the tea ridge accurately and navigate the teaplucking machine, and also can solve the disadvantages of cutting leaves with too many old leaves, and lay a solid foundation for fully autocutting type teaplucking machine in the future.