System Of Sugarcane Cutting Devlce To Protect The Bud Based On Computer Vision

Sugarcane is the main raw materials in sugar industry. The growing area mainly distributed in Guangxi, Guangdong, Hainan, Fujian and Yunnan and other provinces. Our country is one of the world largest sugarcane producer. The development of sugar cane planting directly affects thousands of sugar people’s living. Sugarcane planting is a labor-intensive production activities, the processes including ditching, fertilizing, voted species, cover soil compaction, laying film and Pressure film. People planted sugar cane, per person per day to about0.5acres planted, so it is low efficiency, labor-intensive and poor economic. This shows that the development of the mechanization of sugar cane plantations in the realization of the process of sugar cane fullness mechanization is particularly important.The significance of this paper research is to save cane seeds by cutting down the injured buds rate of septate internodes of the cane seed, to save labor force, cut down the cost of sugar cane production and improve the accuracy of sugar cane mechanised planting. Faced with the current large area of sugar cane planting, the low degree of mechanised planting and the high rate of injured buds in automation planting procedure, this paper bases on MATLAB as a development platform, proposes a new method to identify sugar cane internodes, The main contents include the following points:1. Analysis of sugarcane internode characteristics, and then the cane diameter, the cane internodes width and cane internodes spacing were measured.2. Analysis of the working process of the various components of the the injury prevention buds apparatus and the cooperation between, Motion displacement of the cane in the cutting box, Obtained the camera shutter trigger interval formula in different movement speeds.3. Compared to vertical and vertical-level Sobel edge detection in the edges of the image processing that the level Sobel edge detection is better, the internode can be clearly identified, and eliminated the effects of the sugarcane outside edge and the straight line4. using a method of maximum radius of neighborhood to get a structure element and then to expansion of the target area. After that, The target area is strengthene(?) but not so exaggerated. Provided the ideal signal to the subsequent corrosion and expansion processing.5. Used Radon transform to detect the sugarcane internode edge straight line. Accurately locate the position of the the sugar cane internode in this paper:Blacl cane, whenmax(Rα)=42.8505, x’=0, α=4°; sugarcane, when max(Rα)=75.2418, x’=-12, α=3°;white cane, whenmax(Rα)=90.7042, x’=-2, α=4°. Provides a theoretical basis for the research of the device of anti-bud injury in seedcane cutting.6. Experimental verification of the algorithm. The percent of accurate identify the sugarcane is100%, and the the sugarcane internode of position accurate to0.1mm. Error date is small and Credibility is very high.