SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World Images
Specular highlights detection and removal in images is a fundamental yet non-trivial problem of interest. Most modern techniques proposed are inadequate at dealing with real-world images taken under uncontrolled conditions with the presence of complex textures, multiple objects, and bright colours,...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/17/6552 |
| _version_ | 1827664976412672000 |
|---|---|
| author | Atif Anwer Samia Ainouz Mohamad Naufal Mohamad Saad Syed Saad Azhar Ali Fabrice Meriaudeau |
| author_facet | Atif Anwer Samia Ainouz Mohamad Naufal Mohamad Saad Syed Saad Azhar Ali Fabrice Meriaudeau |
| author_sort | Atif Anwer |
| collection | DOAJ |
| description | Specular highlights detection and removal in images is a fundamental yet non-trivial problem of interest. Most modern techniques proposed are inadequate at dealing with real-world images taken under uncontrolled conditions with the presence of complex textures, multiple objects, and bright colours, resulting in reduced accuracy and false positives. To detect specular pixels in a wide variety of real-world images independent of the number, colour, or type of illuminating source, we propose an efficient Specular Segmentation (SpecSeg) network based on the U-net architecture that is expeditious to train on nominal-sized datasets. The proposed network can detect pixels strongly affected by specular highlights with a high degree of precision, as shown by comparison with the state-of-the-art methods. The technique proposed is trained on publicly available datasets and tested using a large selection of real-world images with highly encouraging results. |
| first_indexed | 2024-03-10T01:15:08Z |
| format | Article |
| id | doaj.art-60c7de7b398d433e965be6800166e4c0 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T01:15:08Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-60c7de7b398d433e965be6800166e4c02023-11-23T14:10:26ZengMDPI AGSensors1424-82202022-08-012217655210.3390/s22176552SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World ImagesAtif Anwer0Samia Ainouz1Mohamad Naufal Mohamad Saad2Syed Saad Azhar Ali3Fabrice Meriaudeau4Laboratoire d’Informatique, du Traitement de l’Information et des Systèmes (LITIS), Normandie Université, UNIROUEN, UNIHAVRE, INSA Rouen, 76000 Rouen, FranceLaboratoire d’Informatique, du Traitement de l’Information et des Systèmes (LITIS), Normandie Université, UNIROUEN, UNIHAVRE, INSA Rouen, 76000 Rouen, FranceCentre for Intelligent Signal& Imaging Research (CISIR), Electrical and Electronic Engineering Department, Universiti Teknologi PETRONAS, Seri Iskandar 32610, MalaysiaAerospace Engineering Department, King Fahd University of Petroleum & Minerals, Dhahran 31261, Saudi ArabiaImViA, Université Bourgogne-Franche-Comté, 71200 Le Creusot, FranceSpecular highlights detection and removal in images is a fundamental yet non-trivial problem of interest. Most modern techniques proposed are inadequate at dealing with real-world images taken under uncontrolled conditions with the presence of complex textures, multiple objects, and bright colours, resulting in reduced accuracy and false positives. To detect specular pixels in a wide variety of real-world images independent of the number, colour, or type of illuminating source, we propose an efficient Specular Segmentation (SpecSeg) network based on the U-net architecture that is expeditious to train on nominal-sized datasets. The proposed network can detect pixels strongly affected by specular highlights with a high degree of precision, as shown by comparison with the state-of-the-art methods. The technique proposed is trained on publicly available datasets and tested using a large selection of real-world images with highly encouraging results.https://www.mdpi.com/1424-8220/22/17/6552specular highlightsimage segmentation |
| spellingShingle | Atif Anwer Samia Ainouz Mohamad Naufal Mohamad Saad Syed Saad Azhar Ali Fabrice Meriaudeau SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World Images Sensors specular highlights image segmentation |
| title | SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World Images |
| title_full | SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World Images |
| title_fullStr | SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World Images |
| title_full_unstemmed | SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World Images |
| title_short | SpecSeg Network for Specular Highlight Detection and Segmentation in Real-World Images |
| title_sort | specseg network for specular highlight detection and segmentation in real world images |
| topic | specular highlights image segmentation |
| url | https://www.mdpi.com/1424-8220/22/17/6552 |
| work_keys_str_mv | AT atifanwer specsegnetworkforspecularhighlightdetectionandsegmentationinrealworldimages AT samiaainouz specsegnetworkforspecularhighlightdetectionandsegmentationinrealworldimages AT mohamadnaufalmohamadsaad specsegnetworkforspecularhighlightdetectionandsegmentationinrealworldimages AT syedsaadazharali specsegnetworkforspecularhighlightdetectionandsegmentationinrealworldimages AT fabricemeriaudeau specsegnetworkforspecularhighlightdetectionandsegmentationinrealworldimages |