Digital Image Correlation Method Based on SURF in Airship Envelope Measurement

Abstract

Digital image correlation (DIC) based on computer vision is introduced to the measurement of airship envelope because the material has the characteristics of small modulus and large strain range which bring some difficulties to the traditional contact methods. This paper presents a proposed DIC initial value estimation algorithm by introducing SURF (Speeded Up Robust Feature) method to improve the correlation and system automation, which makes the instability and the initial value sensitivity problems of further optimization algorithm solved. Firstly, SURF is used to detect the feature points in the images. Then, the SURF feature points are associated with sub-regions by selecting N feature points, which means that per sub-region is represented by few feature points. Finally, the selected points in the images before and after the deformation are accurately matched so that displacement and strain initial values are obtained for Newton-Raphson (N-R) method. The algorithm is verified by simulated speckle experiment and airship envelope tensile experiment and is compared with the traditional initial value estimation method. The results show that the proposed algorithm can provide more stable and accurate initial value estimation which not only reduces operational complexity but also speeds up the matching process.