Multiple-View Matching AMMGC Model for Three-Line-Array Digital Images

In digital photogrammetry, single stereo image matching models, which may obtain multiple solutions, have low matching reliability and successful rate. Therefore, subsequent productions are greatly constrained. Based on object-space-based ADS40 multiple-image matching models, image-space-based AMMGC multiple-image matching model is proposed. AMMGC model is appropriate for the matching of three-line-array digital images, synthesizes all the information of matching images to avoid blind information area, and obtains a single solution. It checks matching results by image checking strategy and optimizes image matching results by adaptive window extension strategy. Experimental results prove that AMMGC model can obtain reliable matching results for similar feature points and discontinuous feature points, and thus, possesses greatly practicability.