Akademik Veri Yönetim Sistemi
1st International Congress of Electrical and Computer Engineering, ICECENG 2022, Bandirma, Türkiye, 9 - 12 Şubat 2022, cilt.436 LNICST, ss.57-67
The images which are captured in indoors and/or outdoors may be badly impacted when sufficient light does not exist. The pictures’ low dynamic range and high noise levels may have an impact on the overall success of computer vision systems. Computer vision applications become more powerful in low light situations when low light picture augmentation approaches are used to boost image visibility. Low light photos have a histogram that is very similar to hazy photographs. As a result, haze reduction techniques can be utilized to increase low light photo contrast. An image improvement approach based on inverting low lighting images and applying picture dehazing with an atmospheric light scattering model is suggested in this paper. The suggested technique has been implemented on the Android operating system. The proposed method delivers about 3 frames per second for 360p video on the Android operating system. It is extremely feasible to increase this real-time performance by employing more powerful hardware.