Research On Placement Head Positioning And Chip Component Correction For Surface Mount System Of Based On DSP

Machine vision systems are one of the most important parts of the placement machine. The main function is to locate and correct components of the PCB board. In view of the traditional placement machine is easy to damage components, low precision, slow speed, in this paper, we propose using placement head positioning and correction algorithms based on DSP.This paper studies the machine visual system with DSP as the main processor, the use of CMOS image sensor for image acquisition, combined with USB2.0 for image transmission. The process of research is divided into two parts, the first part propose the location of the PCB based on phase consistency and Hough circle detection, to complete the image median filtering, morphological processing, binarization, phase congruency edge detection processing, using Hough circle detection, calculated the center coordinate which PCB board on the diagonal two mark points in the image coordinate system, by the relationship between the two coordinate systems, which is converted into machine coordinates and calculate the deflection angle of the PCB board. Then one of the mark points as reference point, the use of PCB board in the database information, calculated the actual machine location of each element of the PCB, and uploaded it to the control machine. The second part complete the rectification of the component base on the least square method and grey value to calculate the center of mass, complete the template matching to get the image coordinates of the element side pin, the lease square method is used to carry out the straight line fitting, and the deflection angle is obtained, using gray value weighted method to calculate the coordinate calculation obtained from the central element of the actual deviation distance, then the deflection angle and the offset distance through the USB transmission to the control machine and motion compensation. The experimental results show that the visual system is stable, effective and reliable, and it has a certain reference value of application in the processing of the visual system of the SMT.