Transparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural Network
Abstract The detection and reconstruction of transparent objects have remained challenging due to the absence of their features and variations in the local features with variations in illumination. In this paper, both compressive sensing (CS) and super-resolution convolutional neural network (SRCNN)...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
SpringerOpen
2022-04-01
|
Series: | Photonic Sensors |
Subjects: | |
Online Access: | https://doi.org/10.1007/s13320-022-0653-x |
_version_ | 1811290465653227520 |
---|---|
author | Anumol Mathai Li Mengdi Stephen Lau Ningqun Guo Xin Wang |
author_facet | Anumol Mathai Li Mengdi Stephen Lau Ningqun Guo Xin Wang |
author_sort | Anumol Mathai |
collection | DOAJ |
description | Abstract The detection and reconstruction of transparent objects have remained challenging due to the absence of their features and variations in the local features with variations in illumination. In this paper, both compressive sensing (CS) and super-resolution convolutional neural network (SRCNN) techniques are combined to capture transparent objects. With the proposed method, the transparent object’s details are extracted accurately using a single pixel detector during the surface reconstruction. The resultant images obtained from the experimental setup are low in quality due to speckles and deformations on the object. However, the implemented SRCNN algorithm has obviated the mentioned drawbacks and reconstructed images visually plausibly. The developed algorithm locates the deformities in the resultant images and improves the image quality. Additionally, the inclusion of compressive sensing minimizes the measurements required for reconstruction, thereby reducing image post-processing and hardware requirements during network training. The result obtained indicates that the visual quality of the reconstructed images has increased from a structural similarity index (SSIM) value of 0.2 to 0.53. In this work, we demonstrate the efficiency of the proposed method in imaging and reconstructing transparent objects with the application of a compressive single pixel imaging technique and improving the image quality to a satisfactory level using the SRCNN algorithm. |
first_indexed | 2024-04-13T04:13:21Z |
format | Article |
id | doaj.art-93cf161c056541cba3a748e683529333 |
institution | Directory Open Access Journal |
issn | 1674-9251 2190-7439 |
language | English |
last_indexed | 2024-04-13T04:13:21Z |
publishDate | 2022-04-01 |
publisher | SpringerOpen |
record_format | Article |
series | Photonic Sensors |
spelling | doaj.art-93cf161c056541cba3a748e6835293332022-12-22T03:03:03ZengSpringerOpenPhotonic Sensors1674-92512190-74392022-04-0112411210.1007/s13320-022-0653-xTransparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural NetworkAnumol Mathai0Li Mengdi1Stephen Lau2Ningqun Guo3Xin Wang4School of Engineering, Monash University MalaysiaSchool of Engineering, Monash University MalaysiaSchool of Engineering, Monash University MalaysiaSchool of Engineering, Monash University MalaysiaSchool of Engineering, Monash University MalaysiaAbstract The detection and reconstruction of transparent objects have remained challenging due to the absence of their features and variations in the local features with variations in illumination. In this paper, both compressive sensing (CS) and super-resolution convolutional neural network (SRCNN) techniques are combined to capture transparent objects. With the proposed method, the transparent object’s details are extracted accurately using a single pixel detector during the surface reconstruction. The resultant images obtained from the experimental setup are low in quality due to speckles and deformations on the object. However, the implemented SRCNN algorithm has obviated the mentioned drawbacks and reconstructed images visually plausibly. The developed algorithm locates the deformities in the resultant images and improves the image quality. Additionally, the inclusion of compressive sensing minimizes the measurements required for reconstruction, thereby reducing image post-processing and hardware requirements during network training. The result obtained indicates that the visual quality of the reconstructed images has increased from a structural similarity index (SSIM) value of 0.2 to 0.53. In this work, we demonstrate the efficiency of the proposed method in imaging and reconstructing transparent objects with the application of a compressive single pixel imaging technique and improving the image quality to a satisfactory level using the SRCNN algorithm.https://doi.org/10.1007/s13320-022-0653-xTransparent object imagingsingle-pixel imagingcompressive sensingtotal-variation minimizationSRCNN algorithm |
spellingShingle | Anumol Mathai Li Mengdi Stephen Lau Ningqun Guo Xin Wang Transparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural Network Photonic Sensors Transparent object imaging single-pixel imaging compressive sensing total-variation minimization SRCNN algorithm |
title | Transparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural Network |
title_full | Transparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural Network |
title_fullStr | Transparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural Network |
title_full_unstemmed | Transparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural Network |
title_short | Transparent Object Reconstruction Based on Compressive Sensing and Super-Resolution Convolutional Neural Network |
title_sort | transparent object reconstruction based on compressive sensing and super resolution convolutional neural network |
topic | Transparent object imaging single-pixel imaging compressive sensing total-variation minimization SRCNN algorithm |
url | https://doi.org/10.1007/s13320-022-0653-x |
work_keys_str_mv | AT anumolmathai transparentobjectreconstructionbasedoncompressivesensingandsuperresolutionconvolutionalneuralnetwork AT limengdi transparentobjectreconstructionbasedoncompressivesensingandsuperresolutionconvolutionalneuralnetwork AT stephenlau transparentobjectreconstructionbasedoncompressivesensingandsuperresolutionconvolutionalneuralnetwork AT ningqunguo transparentobjectreconstructionbasedoncompressivesensingandsuperresolutionconvolutionalneuralnetwork AT xinwang transparentobjectreconstructionbasedoncompressivesensingandsuperresolutionconvolutionalneuralnetwork |