Fig. 1. Experimental setup of off-axis DH for aperture synthesis.

Fig. 2. Procedure to verify the effectiveness of the DID indicator.

Fig. 3. Relationship between the DID indicator and the assumed registration error.

Fig. 4. Comparison of synthesized phase images (the center part with 1280 × 1280 pixels) by different methods. (a)–(c) are the synthesized 3D phase images by the state-of-the-art method with SSD, CC, and ZNSSD criteria, respectively; (d), (e) are the synthesized 3D phase images by the proposed method with ZNSSD and SSD criteria, respectively; (f) is the synthesized 3D phase image by traversing all of the positions for the optimal registration positions with the maximum DID value.

Fig. 5. Comparison of square dotted line areas in Figs. 4(c) and 4(e) by different methods. (a) is the best result of the state-of-the-art method (with ZNCC criterion), (b) is the best result of the proposed method (with SSD criterion), (c) comparison of profiles of (a) and (b) in their 450th rows (marked with dotted line).