Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors
The CMOS (Complementary Metal-Oxide-Semiconductor) is a new type of solid image sensor device widely used in object tracking, object recognition, intelligent navigation fields, and so on. However, images captured by outdoor CMOS sensor devices are usually affected by suspended atmospheric particles...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2017-09-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/17/10/2175 |
| _version_ | 1797999491351052288 |
|---|---|
| author | Chen Qu Du-Yan Bi Ping Sui Ai-Nong Chao Yun-Fei Wang |
| author_facet | Chen Qu Du-Yan Bi Ping Sui Ai-Nong Chao Yun-Fei Wang |
| author_sort | Chen Qu |
| collection | DOAJ |
| description | The CMOS (Complementary Metal-Oxide-Semiconductor) is a new type of solid image sensor device widely used in object tracking, object recognition, intelligent navigation fields, and so on. However, images captured by outdoor CMOS sensor devices are usually affected by suspended atmospheric particles (such as haze), causing a reduction in image contrast, color distortion problems, and so on. In view of this, we propose a novel dehazing approach based on a local consistent Markov random field (MRF) framework. The neighboring clique in traditional MRF is extended to the non-neighboring clique, which is defined on local consistent blocks based on two clues, where both the atmospheric light and transmission map satisfy the character of local consistency. In this framework, our model can strengthen the restriction of the whole image while incorporating more sophisticated statistical priors, resulting in more expressive power of modeling, thus, solving inadequate detail recovery effectively and alleviating color distortion. Moreover, the local consistent MRF framework can obtain details while maintaining better results for dehazing, which effectively improves the image quality captured by the CMOS image sensor. Experimental results verified that the method proposed has the combined advantages of detail recovery and color preservation. |
| first_indexed | 2024-04-11T11:05:33Z |
| format | Article |
| id | doaj.art-afe90819d45146baa6d5190c3d050492 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-11T11:05:33Z |
| publishDate | 2017-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-afe90819d45146baa6d5190c3d0504922022-12-22T04:28:24ZengMDPI AGSensors1424-82202017-09-011710217510.3390/s17102175s17102175Robust Dehaze Algorithm for Degraded Image of CMOS Image SensorsChen Qu0Du-Yan Bi1Ping Sui2Ai-Nong Chao3Yun-Fei Wang4College of Aeronautics and Astronautics Engineering, Air Force Engineering University, Xi’an 710038, ChinaCollege of Aeronautics and Astronautics Engineering, Air Force Engineering University, Xi’an 710038, ChinaInformation and Navigation College, Air Force Engineering University, Xi’an 710077, ChinaCollege of Aeronautics and Astronautics Engineering, Air Force Engineering University, Xi’an 710038, ChinaCollege of Aeronautics and Astronautics Engineering, Air Force Engineering University, Xi’an 710038, ChinaThe CMOS (Complementary Metal-Oxide-Semiconductor) is a new type of solid image sensor device widely used in object tracking, object recognition, intelligent navigation fields, and so on. However, images captured by outdoor CMOS sensor devices are usually affected by suspended atmospheric particles (such as haze), causing a reduction in image contrast, color distortion problems, and so on. In view of this, we propose a novel dehazing approach based on a local consistent Markov random field (MRF) framework. The neighboring clique in traditional MRF is extended to the non-neighboring clique, which is defined on local consistent blocks based on two clues, where both the atmospheric light and transmission map satisfy the character of local consistency. In this framework, our model can strengthen the restriction of the whole image while incorporating more sophisticated statistical priors, resulting in more expressive power of modeling, thus, solving inadequate detail recovery effectively and alleviating color distortion. Moreover, the local consistent MRF framework can obtain details while maintaining better results for dehazing, which effectively improves the image quality captured by the CMOS image sensor. Experimental results verified that the method proposed has the combined advantages of detail recovery and color preservation.https://www.mdpi.com/1424-8220/17/10/2175CMOS image sensorsimage dehazeatmospheric scattering modellocal consistent Markov random field |
| spellingShingle | Chen Qu Du-Yan Bi Ping Sui Ai-Nong Chao Yun-Fei Wang Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors Sensors CMOS image sensors image dehaze atmospheric scattering model local consistent Markov random field |
| title | Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors |
| title_full | Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors |
| title_fullStr | Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors |
| title_full_unstemmed | Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors |
| title_short | Robust Dehaze Algorithm for Degraded Image of CMOS Image Sensors |
| title_sort | robust dehaze algorithm for degraded image of cmos image sensors |
| topic | CMOS image sensors image dehaze atmospheric scattering model local consistent Markov random field |
| url | https://www.mdpi.com/1424-8220/17/10/2175 |
| work_keys_str_mv | AT chenqu robustdehazealgorithmfordegradedimageofcmosimagesensors AT duyanbi robustdehazealgorithmfordegradedimageofcmosimagesensors AT pingsui robustdehazealgorithmfordegradedimageofcmosimagesensors AT ainongchao robustdehazealgorithmfordegradedimageofcmosimagesensors AT yunfeiwang robustdehazealgorithmfordegradedimageofcmosimagesensors |