Font Size: a A A

Autonomous Research Of Agricultural Robot Based On Machine Vision

Posted on:2015-07-09Degree:MasterType:Thesis
Country:ChinaCandidate:Q WanFull Text:PDF
GTID:2323330482970778Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Since the rapid penetration of artificial intelligence to the field of agriculture from the end of the 20th century, countries in the world have begun to research the technology of agricultural robotics. China has begun to research the technology of agricultural robotics in the 1990s, and we have achieved great progress. But due to several reasons, this technique has not been applied to the field of agriculture very well. First, some key technologies of agricultural robots still need further research and improvement. Second, the manufacturing and maintenance of agricultural robots cost a lot. With the gradually increasing trend of China's urbanization and population aging, the population engaged in agricultural production is gradually decreasing, the cost increases in agricultural production make the price of agricultural products increase. So it seems to be quite important to develop an agricultural robot based on autonomous operation which is successful for its low-cost, adaptability and multifunction. The main contents of this paper are:This paper designed a multifunctional autonomous system that can be used in the agricultural robot. (Including the spraying operations on vegetables, spraying herbicides on weeds and mechanical weeding operations on weeds), this system will be applied to the agricultural robot, and then an agricultural robot with a variety of autonomous functions has been designed. This robot not only can improve the functions of agricultural robot and promote its development and application, but it also can largely save the use of chemical pesticides and reduce the pollution through the spraying operations and mechanical weeding operations of precise positioning for vegetables and weeds.1. The agricultural robot's autonomous operation system has been planed generally. This autonomous agricultural robot was mainly used the machine vision to realize autonomous navigation and autonomous operation, this paper focuses on the research of the autonomous operation system and then the design objective and layout arrangement of each part was been determined. This paper focuses on the working principle of the system.2. The design of the device system of operating tools. This paper conducted the design work for working device based on the feature of the working target and working manner of agricultural robot. The design work was mainly split into spraying module and mechanical weeding module. What's more, the carrier of working tools was been designed already which uses the platform of X-Z as the carrier3. The research of the image processing algorithm. Here it mainly gave an account of how to use the machine vision system to recognize weeds and vegetables. Using elongation and other shape feature parameters to accomplish the separation and identification of weeds and vegetables, and then detail discourses and experimental verification were done for this algorithm. The information about the coordinate positions of the weeds and vegetables were obtained finally.4. The control system was designed and programmed. This coordinate information which calculated by the host computer is transmitted via serial communication to MCU, by using an appropriate control program of MCU to control the motion of stepper motor and agricultural robot. Then our working tool was being driven to reach a specified target location. And then the working motor was controlled by an on-off operation of relay and solenoid valves to complete the appropriate work. After the work was completed, the stepper motor was controlled to drive the operational tools to return to the job center position follow original way.5. Test and results analysis were conducted. The speed of agricultural robot is 0.1-0.5m/s in parking mode and traveling mode, five groups of tests were conducted on vegetables spraying operations, spraying herbicides on weeds and mechanical operations weeding respectively. A detailed statistics and analysis was adopted for the accuracy rate and time of the operation. At last, the best operating mode and speed of crop spraying operations, spraying herbicides on weeds and mechanical operations weeding operations were confirmed.
Keywords/Search Tags:Agricultural robot, Autonomous operation, Machine vision, Image processing, Weed identification
PDF Full Text Request
Related items