Cnc System Based On Matlab Image Processing Module

With the fast development of the computer software and hardware, computer numerical control system (CNC) has become a main product of numerical control system, and it also has played an important role in machinery manufacture. For a CNC system, pretreatment and feature analysis of an image are very important in order to get a better manufacturing level, canonical craftwork and a strict precision criterion. The success of numerical control tools depends on the correct data that can be sending to the control center at the beginning.In this thesis, the research is mainly about follows: Firstly, make pretreatment on the image supposed to be engraved, such as graying, noise removing, smoothing, enhancing; Secondly, pick up the vector contour of the image in XY plane with the different grey feature of different pixel; Thirdly, obtain the pixel depth in z direction with the pixel grey level; Finally, submit the data including vector contour and the pixel depth to control center. The creativity of this paper lies in two points:1 We proposed a new method based on Orthogonal Polynomial fitting for getting the curve contour. It can correctly pick up the straight-line or curve contour of the original image and fit it with straight-line or curve. The result of experimentation indicated that the effect achieved by this method was much better than by simplestraight line fitting in the case of little increase of compute and memory. 2 We inducted"Generalized Fuzzy Operator (GFO)"conception. Preprocessed the image block one by one with GFO, and obtained the pixel depth in z direction with grey-depth mapping. The result of experimentation indicated that the 3d display of the original image has gotten an obvious improvement.3 Using Mat lab and VC++6.0 as tools, I made the module of processing the gray images and the result can make good.