Three Translations Parallel Transplant Robot And It's Machine Vision System Research

High-efficient factory agricultural production is a domestic and overseas new developing industry technology, which has already moved towards practicality and industrialization in some developed countries. Because the workforce is with high costs in developed countries, they have already popularized the mechanized operation in the greenhouse. The application of the robot in the plant factory has already got higher attention, and it has made the preliminary achievement, however, the research in this respect is still relatively deficient in our country. Parallel mechanisms have many advantages over serial mechanisms in terms of high stiffness, precision and low inertia, furthermore, they are easily controlled, and which have remedied the disadvantages of serial mechanisms. Considered the situation mentioned above, a three translations parallel robot is introduced to the automatic transplanting system, and the visual system of the machine was added to the system, which made the translations robot more flexible in autoimmunization.The paper carries on the design, model simulation on the basis of 3 - CRR model machine, improve partial structure was improved, besides ,the size, the method of installing, transmission system and manipulator were designed . The right parts of an apparatus were chosen and a new type of three translations parallel robot which was full decoupled was produced.CMOS transducer camera of the low-priced USB interface was used as the vision to gather the apparatus, which reduced the systematic hardware cost of the vision and simplified the structure of the visual system, and is favorable to the machine vision popularization and application in agriculture. Through studying the relevant knowledge of machine vision, I understand collection, cutting apart, dealing with of the image and camera demarcating of the camera. Study the segmentation method for color image and use HSI color law and ultra green to cut apart the image of seedlings. First, study colored method of cutting apart of picture, use HSI color law with ultra to cut apart method cut apart the seedling picture separately. Then the picture after cutting apart carries on two value treatment. In the following, use the morphological method of number value to get the position of centre of every seedling in the image. Finally choose ultra green to cut apart on the base of simple procedure, little running time, and the centre position got to deal with closed to the seedling actual rhizome position. Understand the standard relevant knowledge of the camera, Carry on the parameter of the camera according to the standard method of camera based on that the radial restrains that Tsai puts forward. Write procedure to carry on pattern process and change world three-dimensional coordinate into the coordinate of every seedling picture element coordinate of the center, So that the executive body can move the moving platform of the manipulator above the seedling and pick the seeding accurately. Compared the coordinate computed by machine vision system with the actual position of the seedlings, and analyze the error.