Research On On-line Visual Identification And Location Of Typical Package Chip Surface Mount

As a carrier of integrated circuit,the chip can make it a useful microelectronic component by surface mounting.It has been widely applied in the field of civil and military industry.With the continuous improvement of the performance index of chip mounting,had carried out in-depth research on surface mounting technology,analyzed the advantages and disadvantages of the existing surface mounting technology,the online recognition and positioning scheme for chip surface mount based on machine vision is proposed,and demonstrated the feasibility of the scheme.Based on the principle of machine vision system,the dissertation studies the experimental system of online visual identification and location on chip surface and introduces he principle and calibration method of industrial camera as well as the type and selection method of the light source.The core technology of vision system is image processing algorithm.By selecting different types of typical chips,the thesis designs a reasonable visual recognition and location algorithm,and analyzes the feasibility of the algorithm based on simulation experiments.In the process,the thesis studies the filtering de-noising,threshold segmentation,feature extraction and matching algorithms,to realize intelligent recognition and location of typical chip's type,number,pin spacing and height.The main contents of the article are as follows:(1)The dissertation studies and analyzes the basic theory of machine vision technology,and choose a suitable camera and light source to set up an experimental platform for visual identification and positioning.At the same time,the camera calibration is carried out by using the Zhang's demarcation method.The visual system selects different modes of light according to the characteristics of different chips when collecting images,and sampling and photographing several kind of chips for research in the appropriate light environment.(2)The thesis studies the methods of on-line identification and location of the chip,and analyzes the production demand of chip surface mounting.The dissertation also points out the advantages and disadvantages of the chip recognition and positioning technology,puts forward an online recognition and positioning scheme for chip surface mount based on machine vision,and demonstrates the advantages and feasibility of the scheme.The key technology in the scheme is the image processing technology,so it studies a variety of image processing algorithms.(3)In view of the research on online visual recognition and positioning of chip surface mount technology,The thesis selects SIM chip,QFP chip and BGA chip for simulation and analysis.In order to eliminate the interference points in the chip image,The target area filtering algorithm is used to eliminate the points;in order to identify the different types of chips,based on the appearance features of the chip,the thesis a variety of different recognition algorithms,such as invariant moments,rectangles and template matching.The experimental results show that the rectangle algorithm is easier to identify the target area of the approximate rectangle in the chip image.The template matching algorithm and the invariant moment algorithm can also identify the target area in the different chip images.But the complexity is relatively high and time consuming is relatively more.(4)Aim at the study on the precise positioning of the rectangular chip,in order to obtain high-precision position coordinates and angle of deflection,the thesis extracting the edge points of the rectangular chip,the distance between the center and the edge point is used to determine the four vertices.Finally,the edge points are respectively fitted by four vertices,obtains rectangular contour of the chip,and extracts the pose of chip.In summary,the experiment shows that the algorithm proposed in this thesis is feasible.The algorithm has better accuracy,stability and faster speed.Can be realized on the chip surface mount online identification and location of demand,to accomplish the task of identifying and locating the chip.