Low-Illuminance Texture Image Enhancement Method Based on SCBSO Algorithm

Fig. 2. Schematics of chromosome structure before and after mapping. (a) Distribution of gray value of original image; (b) gray value distribution of remapped image; (c) original chromosome structure; (d) remapped chromosome structure

Download full size

Fig. 3. Flowchart of SCBSO algorithm

Download full size

Fig. 4. Image enhancement method for simple chromosome structure of SCBSO

Download full size

Fig. 5. Iterative curves of four algorithms for f₁(x) function

Download full size

Fig. 6. Iterative curves of four algorithms for f₂(x) function

Download full size

Fig. 7. Image collected in database

Download full size

Fig. 8. Image extracted in ROI

Download full size

Fig. 9. Experimental image 1 enhanced by different algorithms. (a) Enhanced image of original image; (b) image enhanced by BSO; (c) image enhanced by GA; (d) image enhanced by SCBSO

Download full size

Fig. 10. Gray histograms of images enhanced by different algorithms. (a) Gray histogram of original image; (b) gray histogram enhanced by BSO; (c) gray histogram enhanced by GA; (d) gray histogram enhanced by SCBSO

Download full size

Fig. 11. Experimental image 2 enhanced by different algorithms. (a) Original image; (b) image enhanced by BSO; (c) image enhanced by GA; (d) image enhanced by SCBSO

Download full size

Fig. 12. Gray histograms of image 2 enhanced by different algorithms. (a) Gray histogram of original image; (b) gray histogram enhanced by BSO; (c) gray histogram enhanced by GA; (d) gray histogram enhanced by SCBSO

Download full size

Table 1. Results of benchmark function test
View table
Table 1. Results of benchmark function test
Function Dimension ofindependent variable Range ofindependent variables Function minimum
f₁(x)= $\begin{matrix} \overset{n}{\sum_{i = 1}} \end{matrix} \begin{matrix} |x_{i}| \end{matrix}$ + $\begin{matrix} \prod_{i}^{=} \end{matrix} \begin{matrix} |x_{i}| \end{matrix}$ 30 [-10,10] 0
f₂(x)= $\begin{matrix} \overset{n}{\sum_{i = 1}} \end{matrix}$ -x_isin( $\begin{matrix} \sqrt[]{|x_{i}|} \end{matrix}$ ) 30 [-50,50] -418.9829×n

Table 2. Comparison of function performances of different algorithms

View table

Table 2. Comparison of function performances of different algorithms

Function	Algorithm	Function mean	Variance	Time /s
	SCBSO	0	0	0.398
f₁(x)= $\begin{matrix} \overset{n}{\sum_{i = 1}} \end{matrix} \begin{matrix} \|x_{i}\| \end{matrix}$ + $\begin{matrix} \prod_{i}^{=} \end{matrix} \begin{matrix} \|x_{i}\| \end{matrix}$	BSO	1.130×10^-4	3.5700×10^-4	0.477
	PSO	1.225	3.2740	0.401
	GA	0.005	0.0032	1.127
	SCBSO	-12569	5.4210	0.421
f₂(x)= $\begin{matrix} \overset{n}{\sum_{i = 1}} \end{matrix}$ -x_isin( $\begin{matrix} \sqrt[]{\|x_{i}\|} \end{matrix}$ )	BSO	-12214	271.4000	0.494
	PSO	-10989	624.7000	0.417
	GA	-9984	2464.2000	1.629

Table 3. Objective evaluation index of images enhanced by different algorithms
View table
Table 3. Objective evaluation index of images enhanced by different algorithms
Image Method LOE VIF PSNR
SCBSO 37.124 1.1074 18.662
Image1 BSO 70.487 0.7219 16.096
GA 45.815 1.0119 17.486
SCBSO 30.223 1.2471 20.095
Image2 BSO 39.912 1.0891 18.781
GA 38.351 1.1299 19.813

Table 4. Average indexes of different methods on 40 images
View table
Table 4. Average indexes of different methods on 40 images
Image Method LOE VIF PSNR
SCBSO 40.925 1.1219 19.742
Database 1 BSO 65.797 0.9733 17.844
GA 47.584 1.0174 18.978
SCBSO 31.595 1.2146 20.846
Database 2 BSO 40.362 1.0941 19.461
GA 39.432 1.1764 18.456

Tools

Get Citation

Copy Citation Text

Zhiyong Tao, Lei Zhang, Sen Lin. Low-Illuminance Texture Image Enhancement Method Based on SCBSO Algorithm[J]. Laser & Optoelectronics Progress, 2019, 56(24): 241002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites