Key Technologies Of Vegetable Plug Seedlings Transplanting Based On Machine Vision And Simulation Test

At present, the mainly method for vegetable plug seedlings transplanting is manual in domestic. As the important link of breeding process, the realization of automation and mechanization of plug seedlings transplanting equipments could reduce the intensity of labor, improve the efficiency of working, and ensure the consistency of seedlings. Following the maturity and popularization, the backward of equipments for automatic plug seedlings transplant become the biggest obstacle, no applicable automatic equipments could be used in actual production. When we tried to use transplanting equipments to transplant the plug seedlings, the fracture of plug seedlings nutritional bowls and the damages of plug seedlings lead to a low success rate of transplanting. The main reason is we don’t have sufficient and overall researches on basic mechanical theories of transplanting, leading to the lack of practical technical support. Many achievements are still in the testing stage, so the development of transplanting machines is restricted. In order to protect the integrity of plug seedlings nutritional bowls during the transplantation. In this paper we study on the key technologies of plug seedlings transplanting based on machine vision and simulation test, in order to improve the efficiency of transplanting, ensure transplanting effects and achieve consistency transplanting.Firstly, we propose the locating method of plug seedlings based on binocular vision to solve the problem that location of plug inaccurate lead to transplanting effect is not ideal. From the perspective of two-dimension and three-dimension respectively, we study on the plug location methods in the aspects of plane and three-dimensional, completed the locations of center point coordinates, angle, and simulated the plug surface shape. So as to achieve the purpose of accurately guide the ending effectors transplant seedlings accurately.Secondly, the nutritional bowl matrixes are easily damaged in the actual process of transplanting, and unified automatic transplanting method can not remedy the damaged nutritional bowl matrixes timely. Thus cause the seedlings after transplanting growth cycle are different, appear uneven, the transplanting effect is not ideal. We proposed recognition of seedlings in mechanical transplanting processing by morphological image detection in order to solve the practical problem above. The method discriminated the degree of nutritional bowl matrixes damages by contrasting the ratios between matrix areas and referenced areas of three angle images from each sample. The purpose of this method is to classify the seedlings transplanted, remedy them in time, and to achieve consistency transplanting.Thirdly, we confirmed that the shape feature of pick-up plug seedlings manipulator structures is the primary factor, and this factor influence the transplanting effect of the nutritional bowl matrixes by testing multi-factor. Testing factors include soil moisture content, shape feature of pick-up plug seedlings manipulator structures, and the performance improvement of root system on soil matrix shear strength at the beginning of crop growth seedlings. We use the soil matrix complete rate as test index, analyzing the relationship between variance and range we have explicated the best range. We adopt orthogonal experiment method to obtain the influence factors primarily, secondary, and the optimal level.Fourthly, we combined the analysis conclusions of multi-factor influenced transplanting test by the ANSYS/LS-DYNA finite element software, and tested the physical parameters of nutritional bowl matrix, meanwhile, we analysis stress response effects that different structures of ending effectors effect on the nutritional bowl matrix.Finally, we carried out overall test, completed the design of test system software by secondary development to achieve the communication with the manipulator control platform. We have designed the overall testing scheme, steps and each test link to verification test on the key technologies proposed in this paper through the gradual and progressive way in order to provide the basics for the feasibility.By studying the key technologies of vegetable plug seedlings transplanting based on machine vision and simulation test in this paper, we draw the following conclusions:(1) From the testing results, it is known that mechanical transplanting plug localization method based on binocular stereo vision can comparatively accurate simulating the surface shape of plug, it has a low overall error, and satisfies the requirements of automatic transplanting application processing.(2) We study the recognition of transplanting nutritional bowls matrix effects by using morphological image detection method. The method is simple and practicable, and we developed a recognize software independently to calculate the soil matrix complete rate to decide if it is qualified. It is shown that our method is short time-consuming and has high accuracy rate, it can match the demand of time on the transplanting line. The method satisfied with the requirement of real-time processing.(3)It is shown that the shape feature of pick-up plug seedlings manipulators structures is the primary factor influence the effect, and determine the u-shape structure is the best. Matrix moisture content has effect on pick-up plug seedlings. The moisture content around 50% is the best. We pick-up plug seedlings when they grow the first pair of leaves, the performance improvement of root system on soil matrix shear strength at the beginning of crop growth seedlings have little effect on soil reinforcement. This test for further study and improve the equipment of pick-up plug seedlings determines the key technical points, so as to establish the important theoretical basis for exploring automatic transplanting for pick-up plug seedlings.(4)Using the ANSYS/LS-DYNA finite element software to simulation analysis, the results are shown that Z-shape and U-shape structures have a reasonable stress distribution when they insert into the nutritional bowl matrix, stress in the nutritional bowl matrix change mildly so it has little disturbance and destruction to the matrix. Doing the simulation analysis of the process three kinds of structures pick-up the nutritional bowl matrix, Z-shape structure has a centered stress, the Von Mises stress inside the nutritional bowl body changes greatly and may makes the nutritional bowl matrix broken. But the Von Mises stress inside B-shape and U-shape structures nutritional bowl matrix change mildly, they have little disturbance and destruction to the matrix, the integrity of nutritional bowl matrix can be maintained during the picking up process. So we can conclude that the ending effectors with U-shape structure is reasonable, the matrix will not be destructed or disturbed, integrity of matrix also can be maintained when we pick it up.(5)The testing results from the comprehensive performance test show that the research methods in our key technologies of vegetable plug seedlings transplanting based on machine vision and simulation test proposed in this paper is accurate, feasible, and each key technology provides technical support for plug seedlings automatic transplanting.From the above, we propose the technical proposals of plug seedlings locations and recognition based on machine vision from the angle of the problems of plug seedlings transplanting. Meanwhile, we have analyzed the influencing factors of soil matrix complete rate, and further simulation analysis of nutritional bowls stress response. Our research can be use as the theoretical basis and technical support for plug seedlings automatic transplanting.