• Use source img + depth data from Deep-AI + homography matrix
• Then stitching the image

As described in [7]

• Step 1: Detecting keypoints (DoG, Harris, etc.) and extracting local invariant descriptors(SIFT, SURF, ORBFASTBRIEF, etc.) from multiple input images
• Step 2: Matching the feature descriptors between the multiple images
• Step 3: Estimating the homography matrixes using the set of matched features of each pair images
• Step 4: The warping transformation image using the homography matrix of each pair images, which are obtained from previous step

• Keypoints
• local invariant descriptors
• image stitching
• homography matrix
  • RANSAC algorithm
• warping transformation

1. image-stitching
2. panoramic-reconstruction
3. image-stitching-deep-learning
4. depth-estimation
5. image-stitching-datasets
6. image-inPainting
7. deep-image-stitching
8. keras-ResNet