Related Experiment Video
Updated: Jun 29, 2026
05:49
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
Occlusion removal method of partially occluded 3D object using sub-image block matching in computational integral
Dong-Hak Shin1, Byung-Gook Lee, Joon-Jae Lee
1Dept of Visual Contents, Dongseo University, Sasang-Gu, Busan, Korea.
Optics Express
|October 15, 2008
Summary
This study introduces a novel occlusion removal technique for 3D object recognition in computational integral imaging (CII). The method enhances 3D image quality by effectively addressing occlusion issues using sub-image block matching.
Area of Science:
- Computer Vision
- Image Processing
- 3D Reconstruction
Background:
- Occlusion significantly degrades the resolution of 3D reconstructed images in computational integral imaging (CII).
- Existing methods struggle to accurately reconstruct 3D objects when partially occluded.
Purpose of the Study:
- To develop and validate an effective occlusion removal method for partially occluded 3D objects in CII.
- To improve the visual quality and recognition accuracy of 3D reconstructed images.
Main Methods:
- Sub-image transform applied to elemental image array (EIA).
- Block matching for depth estimation from sub-images.
- Occlusion removal based on estimated depth information.
- Inverse sub-image transform to obtain a modified EIA.
- Reconstruction of 3D images using the modified EIA.
Main Results:
- Successful removal of occlusion from 3D reconstructed images.
- Substantial gain in the visual quality of 3D images.
- Demonstrated effectiveness through experimental validation.
Conclusions:
- The proposed sub-image block matching method effectively removes occlusion in CII.
- This technique significantly enhances the quality of 3D reconstructed images.
- The method offers a valuable solution for recognizing partially occluded 3D objects.
