Adaptive Underwater Image Enhancement Algorithm

Ning Yang; Haibing Su; Tao Zhang

doi:10.3788/LOP231335

Laser & Optoelectronics Progress, Volume. 61, Issue 8, 0837001(2024)

Adaptive Underwater Image Enhancement Algorithm

Ning Yang^1,2,3, Haibing Su^1,2,3,4、*, and Tao Zhang^1,2,4

Author Affiliations

¹Institute of Optics and Electronics, Chinese Academy of Sciences, Chengdu 610209, Sichuan , China

²National Key Laboratory of Optical Field Manipulation Science and Technology, Chinese Academy of Sciences, Chengdu 610209, Sichuan , China

³School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

⁴School of Optoelectronics, University of Chinese Academy of Sciences, Beijing 100049, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(9)

Fig. 1. Algorithm flow chart

Download full size

Fig. 2. Results of adaptive color equalization. (a) Original images; (b) processing results of gray world algorithm; (c) processing results of adaptive color balance

Download full size

Fig. 3. Results of adaptive global contrast enhancement. (a) Original images; (b) processing results of adaptive color equalization; (c) processing results of adaptive global contrast enhancement

Download full size

Fig. 4. Processing results of underwater images in low light. (a) Original images; (b) processing results of adaptive color equalization; (c) processing results of adaptive global contrast enhancement

Download full size

Fig. 5. Results of adaptive underwater image enhancement algorithm. (a) Original images; (b)‒(d) processing results of adaptive color equalization, adaptive global contrast enhancement, and adaptive detail enhancement

Download full size

Fig. 6. Subjective comparison of results using different methods. (a) Original images; (b)‒(g) processing results of UDCP, fusion method, UTV, ULV, L2UWE, and the proposed method

Download full size

Table 1. Objective quality evaluation of Fig. 6

View table

Table 1. Objective quality evaluation of Fig. 6

Metric	Method	image 1	image 2	image 3	image 4	image 5
UCIQE	UDCP^［6］	0.2587	0.2587	0.2587	0.2587	0.5321
	Fusion method^［14］	0.4842	0.5132	0.4922	0.5655	0.4753
	UTV^［8］	0.4024	0.4471	0.5584	0.6078	0.3593
	L2UWE^［13］	0.2587	0.2587	0.2588	0.2594	0.2587
	ULV^［25］	0.4759	0.5345	0.5719	0.6285	0.4817
	Proposed method	0.5882	0.6234	0.6072	0.5941	0.5673
CCF	UDCP^［6］	14.0587	16.7700	16.7085	14.8407	4.5165
	Fusion method^［14］	14.4589	16.9985	17.5129	20.9758	22.3465
	UTV^［8］	47.4826	11.5945	21.9772	45.2255	12.4622
	L2UWE^［13］	16.2607	16.3567	15.2428	18.0635	10.1131
	ULV^［25］	11.0626	16.5197	20.2411	31.4828	21.4070
	Proposed method	26.1962	37.2615	27.8284	45.4160	33.0953
Information entropy	UDCP^［6］	7.1943	7.3265	5.7007	5.6244	6.0584
	Fusion method^［14］	7.4671	7.1428	7.3210	7.1193	7.0869
	UTV^［8］	6.6584	6.5044	5.5432	5.1010	5.8684
	L2UWE^［13］	7.7659	7.4358	7.5794	7.2401	7.5575
	ULV^［25］	7.6663	7.3989	7.8370	6.4850	6.2104
	Proposed method	7.8138	7.7893	7.8355	7.3509	7.6475

Table 2. Average quantitative evaluation of different methods.

View table

Table 2. Average quantitative evaluation of different methods.

Method	UDCP^［6］	Fusion method^［14］	UTV^［8］	L2UWE^［13］	ULV^［25］	Proposed method
UCIQE	0.5973	0.5554	0.5660	0.5590	0.6057	0.6125
CCF	27.4938	22.0796	27.4938	32.6690	27.9457	34.6980
Information entropy	6.7832	7.4114	6.1115	7.5079	7.4537	7.6425

Table 3. Average running time of different methods
View table
Table 3. Average running time of different methods
Method UDCP^［6］ Fusion method^［14］ UTV^［8］ L2UWE^［13］ ULV^［25］ Proposed method
Time 37.557 1.8405 3.5818 45.7899 8.0517 0.3692