Fig. 1. Comparison of detection results of several edge detection operators. (a) Original image; (b) Sobel detection result; (c) Canny detection result; (d) Laplacian detection result

Fig. 2. Gray histograms before and after Laplacian transformation. (a) Interference fringes; (b) gray histograms of speckle background region; (c) gray histograms of speckle interference fringe region

Fig. 3. Segmentation result of second-order gradient entropy

Fig. 4. Result of binary region extraction

Fig. 5. Result of interference fringe region extraction

Fig. 6. Flow chart of algorithm for speckle interference fringe region extraction

Fig. 7. Second-order gradient entropies of different speckle interference fringe patterns. (a) Shearography fringes of circular measured object (left) and their second-order gradient entropy (right); (b) digital speckle interference fringes of circular measured object (left) and their second-order gradient entropy (right); (c) shearography fringes of rectangular measured object (left) and their second-order gradient entropy (right); (d) digital speckle interference fringes of rectangular object with sli

Fig. 8. Gray entropy corresponding to Fig. 7(a)

Fig. 9. Interception results of background and stripe areas

Fig. 10. Verification results of fringe region and background region. (a) Gray entropy of background region; (b) gray entropy of fringe region; (c) second-order gradient entropy of background region; (d) second-order gradient entropy of fringe region

Fig. 11. Classical interference fringe patterns. (a)-(c) Shearography fringes of objects with different shapes; (d)-(e) digital speckle interference fringes of objects with different shapes

Fig. 12. Interference fringe regions manually extracted in classical interference fringes shown in Fig. 11

Fig. 13. Different interference fringe patterns and corresponding extraction results. (a)-(h) Speckle interference fringe patterns; (a')-(h') extraction results of fringe region