Flatness Detection Of Keyboard Caps Based On Structured Light Imaging

Keyboard,as the main input device of electronic equipment,can only be put into the market after passing the test.As one of the major indicators of keyboard qualification,the flatness of key caps also directly affects the comfort and smoothness of users.The traditional key cap flatness detection has some defects in speed or precision,and it is difficult to meet the needs of keyboard manufacturers for online batch detection.Because of the continuous development of linear structured light vision measuring technology and its advantages of non-contact and high-precision,it is of great significance to develop and implement a fast,economical and practical flatness detection device of keyboard caps based on structured light imaging technology.The paper analized the current status of keyboard caps flatness detection and structured light vision measurement at home and abroad.Then,a keyboard caps flatness detection method based on structured light imaging is proposed.Introduced the working principle of the scheme in detail,and expounded the key technologies used in the detection.The main research contents of this paper include:(1)According to the characteristics of the keyboard keys and the requirements of flatness detection accuracy,the relevant system hardware is reasonably selected.Aiming at the contradiction between high resolution imaging and finite field of view.In this work we proposed a multi-camera full-keyboard measurement model of structured light which is based on the principle of structured light triangulation.Finally,the experimental platform of the system has been set up.(2)Since multiple keyboard images are projected with stripes,it is necessary to register and fuse them into a full keyboard image containing all the stripe information.In the light of the characteristics of the system imaging device,this paper narrowed the matching area,and proposed a method of bidirectional feature matching based on epipolar constraint.This method can effectively overcome the mismatching problem caused by one-to-many in unidirectional matching mode.Experimental comparison was conducted with the SURF image Mosaic method,and it further proves the superiority of this algorithm.(3)We designed the keycaps segmentation module to divide the keyboard into mutiple key caps region of interest(ROI),after analyzing the factors that affect the qualit y of optical stripe image of full ke yboard.Then,each keycap region was preprocessed separately to fill the holes inside the stripe and removed the noise and character interference on the keycap.Finally,the fringe region was morphologically trimmed to obtain high-quality light fringe images.We used gray weighted centroid method that based on the skeleton normal direction to extract the midline of stripes.Meanwhile,the thinning method of stripe skeleton is improved.After obtaining the stripe skeleton of single pixel,designed a skeleton pruning algorithm based on weighted longest path to remove the influence of skeleton burrs effectively.In the case of a large number of edge burrs on the edge of the stripe,the sub-pixel position of the stripe center line of the keyframe keycaps can still be extracted accurately and in real time.In this paper,we used the least square method to obtain the plane equation coefficient of standard keyboard caps.According to the distance between the keyboard cap to be measured and the plane where the standard keyboard cap is located,it is used as the criterion to judge the flatness of the keyboard caps.In VS2013 programming environment,completed the realization of each module of the software.Finally,developed a high precision,fast,economical and practical keyboard caps flatness detection device based on structured light imaging.