Huixian Huang, Fanhao Chen. Low-Illumination Image Enhancement Method Based on Attention Mechanism and Retinex[J]. Laser & Optoelectronics Progress, 2020, 57(20): 201004
- Laser & Optoelectronics Progress
- Vol. 57, Issue 20, 201004 (2020)
Fig. 1. Generation process and results of illuminance characteristics. (a) Generation process; (b) illumination feature visualization result
Fig. 2. Add visual results of AM layer. (a) Original images; (b) visualized results
Fig. 3. Results of influence on light map before and after adding attention mechanism layer and its partial enlarged maps.(a) Light maps; (b) reflection maps; (c) light maps without AM layer; (d) light maps with AM layer
Fig. 4. AM-RetinexNet structure diagram
Fig. 5. Y channel frequency diagram of three images
Fig. 6. Comparison of enhancement effect of real indoor low-illumination images and its partial enlarged images. (a) Original images; (b) MSR algorithm; (c) LIME algorithm; (d) LLNet algorithm; (e) RetinexNet algorithm; (f) AM-RetinexNet algorithm; (g) ground truth
Fig. 7. Comparison of enhancement effect of real outdoor low-illumination images and its partial enlarged images. (a) Original images; (b) MSR algorithm; (c) LIME algorithm; (d) LLNet algorithm; (e) RetinexNet algorithm; (f) AM-RetinexNet algorithm; (g) truth map
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Table 1. Comparison results under different imaging parameters and modes
Set citation alerts for the article
Please enter your email address
© Copyright 2018-2021 | Chinese Laser Press. All Rights Reserved 沪ICP备15018463号-20