Study On The Ther Mopaint Auto-recognition Algorithm And The Software Implementation

With the development of modern aerospace,the operating temperature of aviation engine is getting higher, the temperature of the hot components should be detected strictly in order to make the engine parts work effectively and everlastingly, the temperature measurement technology also need to be improve. In numerous temperature measurement technologies, the TSP(Temperature Sensitive Paint) measuring technology has been gradually into the mainstream. However, the temperature recognition is one of the main bottle-neck in restricting the development of this technology. At present stage, the temperature recognition is mainly depended on artificial method, which is inefficient, inaccurate and skilled workers dependence.This paper is dedicated to developing a TSP temperature automatic recognition system to solve the above problem by image processing technology. In this process, the main contents are as the following:Temperature auto-recognition algorithm research includes three parts. First, the isotherm extraction algorithm is considered. In consideration of the characteristics of the thermospaint, the original algorithm was improved and restructured. The new algorithm which combined the optimized K-Means clustering algorithm and canny operator was proposed. The accuracy of isotherm extraction is increased. The second is Image optimization algorithm, the relative research was made in order to improve the accuracy and quality of isotherm. Based on experiments and analysis, the median filter and Gaussian filter have been used to got the final optimization algorithm. The third is temperature recognition algorithm. A new algorithm was proposed in this paper. The isothermal was grouped by high and low threshold comparison method, make full use of the known information of thermos paint, and the isothermal temperature was recognized accurately and efficiently.The system designed of the temperature auto-recognition system. Operators can read temperature easily by following the basic interpretation process which was set up according to the characteristics of temperature auto-recognition system function demand. So it leads to lower requirement for the operator. In addition, human interaction module is added into the process of system designed, the operability and applicability were also improved.According to the characteristics of temperature auto-recognition system function demand, the QT development platform based on C++ and the OpevCV image library were used for the software development, and in the end made the theory study conclusion actually been used in practical.