| Up to now, LED back light has been main light source of LCD, and LED back light module has become the core component of LCD. LED chip and the secondary lens are bonded by plastic. So there will be assembly defects such as center offset, height offset and tilt between them. These defects will affect the uniformity of luminance of lamp beads and the decrease the overall quality of the backlight module. According to the requirements of producing, a kind of detection equipment which can detect the defects in LED back light module will be designed and manufactured.In this paper, we conduct optical modeling and simulation analyzing for single light of different LED back light products, so we know how different lens affect the rays of light source and get the characteristics of the rays and gray curve on the reception screen, at the same time we analyze how the rays distribute after being affected by the lens, then we put forward the corresponding test method. According to different detection method, we derive the algorithm and write the image processing program. The principle of one detection method is based on the changes of difference between right peak and left peak, and the valley value of the gray curve. The gray curve has a certain width of some pixels, passes the central region the image, and has four directions, such as horizontal, vertical.and diagonal. The other one is based on the relationship between image edge and the defection types. Use pixel ellipse fitting to position the image edge coarsely. Use four times polynomial to fit the gray curve along the gradient direction of image edge. Through 5′5 regional gray value interpolation to seek the five adjacent gray value. The gray curve can be obtained by the interpolation and the least squares principle. Take the second derivate of the gray curve and the sub-pixel edge points is the point when the second derivate is zero. Carry on ellipse fitting with these points again to get the accurate image edge. At last calculate the changes of ellipse center coordinates and half shaft length under different defections to detect the defects. Finally, design the mechanical structure of the prototype which includes the overall and the detailed designs. Produce the defect detection device and experiment with different kinds of LED back light products. During the experiments set different defects and get the data. Analyze the data to prove the reliability and stability of the equipment.The detection equipment now can detect defect type of certain products. It can achieve some precision, the speed is fast and the results are reliable. |