Extraction Algorithm Research Of Phase-change Line From Complex Phase-change Thermography Sequence

With the development of aerospace technology, the phase change thermography testing technology, a large area measured temperature technology, was widely applied recently. It was one of the important measurement technology for heat flux distribution lines test on the surface. The key of the technology was the extraction algorithm of phase-change lines and thermal map analysis based on phase-change lines. To achieve the extraction of phase-change lines, this paper made the following work.First of all, based on structural characteristics, an accurate image matching measure was developed to solve the location bias in some frames of phase-change thermography sequence. The matching measure analyzed the features of the original image, counted the edge characteristics, and calculated the weight of the original image in various locations. Then, the max weight value was obtained by comparing all weight of the various values of the picture. According to coordinate transformation to get the accurate initial image, the difference of frame images was used to obtain the effective information of phase change. Due to the unconspicuous difference image,the gray scale contrast method was adopted to enlarge the changes information.Secondly, a phase-change line tracking algorithm was presented by introducing shape priori characteristics into C-V active contour model. Based on the image difference information and the relevance about last image and next image in sequential image frames which called prior knowledge, the computation time and efficiency was improved by this algorithm, which had no repetition reconstruction energy function. The practical of the proposed algorithm was experimented by Matlab, showed that the method was able to rapidly solved multi-phase change lines and topology change problems which could not be solved by existing methods. The effectiveness of the proposed algorithm was validated and compared to alternative methods with variant test models.Finally, to achieve cross-platform development, the algorithm was realized under Windows XP with VC++6.0. Collecting all the phase line into one image, the thermal map of the phase transition was obtained and the change of thermal map in the surface was the heat flow distribution of the test model. The coefficient of aerodynamic heating and thermal coefficient of each model surface position was calculated by the given parameters.The software system based on the algorithm had been successfully applied in practice. By comparing the instance data, the algorithm was proved feasible and reliable.