Research Of Ice Sensor Calibration Based On Image Processing

For a long time people have attached importance to imperilments caused by icing in industry, especially icing on inflight airplanes in field of aviation. In China, We have so unsubstantial technique foundations that it is urgent to engage in promoting research and development of icing sensor. As we know, every sensor must be calibrated before it is used, but the calibration methods which are adopted in our country are still primitive. For example, in some case vernier caliper and even eyes are used for measurement. Therefore, the research and development of calibration system for ice sensors is a vital step in the whole work and playes a very important role.Based on image processing principle, a calibration method for a kind of resonant diaphragm piezoelectric ice sensor is given and researched deeply.The first step of adopting the technology is set up a vision system. The digital camera as one of the key components in the vision system, its performance, especial the accuracy of its calibrated internal parameters, directly affects the measurement precision. Therefore, the camera calibration plays a very important role in image detecte technology. This article analyzes deeply and discusses a model of vision system, and base on it introduces the detailed programme to achieve calibration.The most important part of the thesis is to analyze and process the ice picture based on the image processing algorithms. The work has three step as follows: the first step is to subtract the background from ice picture and select satisfactory filter for the pretreatment of image. Edge extracting is the second step. The thesis compares the differences among various results by adopting different edge extracting algorithms and get a good one from them to find the rough edge. Based on this, we find the fact edge of ice and caculate the ice's thickness according to the camera calibration model.These method is tested on the computer with matlab.