A Deep Learning Framework to Remove the Off-Focused Voxels from the 3D Photons Starved Depth Images
Photons Counted Integral Imaging (PCII) reconstructs 3D scenes with both focused and off-focused voxels. The off-focused portions do not contain or convey any visually valuable information and are therefore redundant. In this work, for the first time, we developed a six-ensembled Deep Neural Network...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Photonics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2304-6732/10/5/583 |
| Summary: | Photons Counted Integral Imaging (PCII) reconstructs 3D scenes with both focused and off-focused voxels. The off-focused portions do not contain or convey any visually valuable information and are therefore redundant. In this work, for the first time, we developed a six-ensembled Deep Neural Network (DNN) to identify and remove the off-focused voxels from both the conventional computational integral imaging and PCII techniques. As a preprocessing step, we used the standard Otsu thresholding technique to remove the obvious and unwanted background. We then used the preprocessed data to train the proposed six ensembled DNNs. The results demonstrate that the proposed methodology can efficiently discard the off-focused points and reconstruct a focused-only 3D scene with an accuracy of 98.57%. |
|---|---|
| ISSN: | 2304-6732 |