Abstract
In recent years, the importance of semantic segmentation has been widely recognized and the field has been actively studied. The existing state-of-the-art segmentation methods show high performance for bright and clear images. However, in low light or nighttime environments, images are blurred and noise increases due to the nature of the camera sensor, which makes it very difficult to perform segmentation for various objects. For this reason, there are few previous studies on multi-class segmentation in low light or nighttime environments. To address this challenge, we propose a modified cycle generative adversarial network (CycleGAN)-based multi-class segmentation method that improves multi-class segmentation performance for low light images. In this study, we used low light databases generated by two road scene open databases that provide segmentation labels, which are the Cambridge-driving labeled video database (CamVid) and Karlsruhe Institute of Technology and Toyota Technological Institute at Chicago (KITTI) database. Consequently, the proposed method showed superior segmentation performance compared with the other state-of-the-art methods.
Highlights
The field of deep-learning-based semantic segmentation has been actively studied since the implementation of the fully convolutional networks (FCN) [1] and SegNet [2] proposed in 2015
It is difficult to train segmentation networks because of lack or inaccurate label information, and the performance improvement is small. In view of these limitations of previous studies, we propose a multi-class segmentation method based on image enhancement using modified CycleGAN in low light or nighttime environments
Unlike existing multi-class segmentation methods in low light or nighttime environments [12]–[15], the modified CycleGAN is used to perform a direct enhancement of low light images to improve the segmentation performance in low light environments
Summary
The field of deep-learning-based semantic segmentation has been actively studied since the implementation of the fully convolutional networks (FCN) [1] and SegNet [2] proposed in 2015. Numerous convolutional neural network (CNN)-based segmentation methods were developed, showing high performance for various segmentation databases. Most semantic segmentation studies mainly handle daytime databases or bright images, and there have been few studies on semantic segmentation dealing with nighttime databases or low light images. Existing methods show good performance for bright and clear images captured in daytime but, the performance drops significantly for nighttime or low light environments. In low light environments, the amount of light is insufficient, and the image is captured with the camera’s exposure time set longer than daytime.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
