| As the key equipment in the nursery factory, the transplanting robot plays an important role in improving the transplanting efficiency. In this paper, the characteristic parameters of plug seedlings were extracted by machine vision technology, the position of seedlings were located and the non-regulated seedlings were identified automatically. Simultaneously, an experimental prototype was designed. The PC control programs which communicated with PLC controller were developed to practice the fixed-point transplant operation of the actuator arm and the end-effector. Experiments show that the accuracy rate of machine vision system was97%and the success rate of transplanting was95.76%, this study will be an example for developing the transplanting robot. Main contents were as follows:1. A suitable machine vision system fit for this research was designed, which was composed of a light source, a CCD camera, a computer and the software of machine vision.2. An appropriate experimental prototype was designed for this research, which was composed of four parts:transmission system, control system, machine vision system and transplanting system.3. Based on the machine vision system, the characteristic parameters of tomato seedlings could be detected, and the vacancy and bad in the plug could be distinguished. The method of contrast adjustment was used to separate image of plug from the transmission belt; The method of image subtraction ExG-ExR was used to separate seedlings from the background of plug; The method of dual fixed threshold was used to extract the plug framework region; The threshold segmentation was got by iterative to obtain the binary image for extracting the parameters of leaf area. Color parameters of R, G and B average value as reference for the bad seedling was counted in pot seedlings. Midpoint method, manual measurement method and centroid method were used to extract the position of seedlings root, and the centroid method was chosen.The result showed the accuracy rate of machine vision system for the format6×12is98.0%comparing with 97.8%for the format8×16.4. PLC control system of transplanting robot was developed. The PLC control system was composed of a PLC controller, a computer, a CCD camera, photoelectric sensors, stepper motors, an electromagnetic relay, and limit switches. By means of R.S-232, computer sent the position of seedlings to pactice motion control. Acceleration and deceleration curve was used to avoid out-of-step and overshoot of the stepper motors. Comparing five different strategies of transplanting, the best strategy for transplanting was chosen to decrease the transplanting time and improve work efficiency.5. The end-effector was developed with fingered grasping, which was droven by stepper motors. Two kinds of structure fingers were compared in the experiment. The result showed that the fingers droven by stepper motors had better performance with the success rate of95.76%, but the fingers with structure of dynamic disk had relatively so large chattering that seedlings slipped off, and the success rate was only52.0%.6. Host and client computer programs were developed. The host computer program, developed by Visual C++6.0, was composed of image acquisition, image processing, serial communication, camera control, feature extraction, position detection, transplant control, image display and user interface. The developing software of client machine is Hollysys Powerpro2.1.2with ladder diagram. The host computer controlled program carried out static analysis and dynamic auto-transplanted control which can be used to many different specifications of plug seedlings. |
PDF Full Text Request