Shadow Compensation of High-Resolution Remote Sensing Images Based on Improved Logarithmic Transformation and Local Enhancement

Yuanyuan Feng; Xianjun Gao; Yuanwei Yang; Fan Deng

doi:10.3788/LOP57.201006

Journals >Laser & Optoelectronics Progress >Volume 57 >Issue 20 >Page 201006 > Article

Laser & Optoelectronics Progress
Vol. 57, Issue 20, 201006 (2020)

Shadow Compensation of High-Resolution Remote Sensing Images Based on Improved Logarithmic Transformation and Local Enhancement

Yuanyuan Feng¹, Xianjun Gao^1、2、*, Yuanwei Yang^1、2, and Fan Deng¹

Author Affiliations

¹School of Geoscience, Yangtze University, Wuhan, Hubei 430100, China

²State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, Hubei 430079, China

show less

DOI: 10.3788/LOP57.201006 Cite this Article Set citation alerts

Yuanyuan Feng, Xianjun Gao, Yuanwei Yang, Fan Deng. Shadow Compensation of High-Resolution Remote Sensing Images Based on Improved Logarithmic Transformation and Local Enhancement[J]. Laser & Optoelectronics Progress, 2020, 57(20): 201006 Copy Citation Text

show less

Fig. 1. Influence of parameter V on logarithmic transformation curve

Download full size

Influence of parameter V on compensation result. (a) Original image; (b) shadow detection result; (c) compensation effect when V = 0.5; (d) compensation effect when V = 1; (e) compensation effect when V=2.5; (f) compensation effect when V=3

Fig. 2. Influence of parameter V on compensation result. (a) Original image; (b) shadow detection result; (c) compensation effect when V = 0.5; (d) compensation effect when V = 1; (e) compensation effect when V=2.5; (f) compensation effect when V=3

Download full size

Fig. 3. Compensation results from three methods. (a) Original image; (b) logarithmic transformation result; (c) local compensation result; (d) joint compensation result

Download full size

Fig. 4. Acquisition diagram of similarity pairs during automatic compensation

Download full size

Fig. 5. Comparison among compensation results of urban image 1. (a) Original image; (b) selected area of detail comparison map; (c) compensation result by Wallis method; (d) local compensation effect; (e) compensation result by original logarithmic transformation; (f) compensation result by proposed method;(g) detail of compensation result by original logarithmic transformation; (h) detail of compensation result by proposed method

Download full size

Fig. 6. Comparison among compensation results of urban image 2. (a) Original image; (b) selected area of detail comparison map; (c) compensation result by Wallis method; (d) local compensation effect;(e) compensation result by original logarithmic transformation; (f) compensation result by proposed method;(g) detail of compensation result by original logarithmic transformation; (h) detail of compensation result by proposed method

Download full size

Fig. 7. Comparison among compensation results of urban image 3. (a) Original image; (b) selected area of detail comparison map; (c) compensation result by Wallis method; (d) local compensation effect;(e) compensation result by original logarithmic transformation; (f) compensation result by proposed method;(g) detail of compensation result by original logarithmic transformation; (h) detail of compensation result by proposed method

Download full size

Fig. 8. Comparison among shadow compensation results of building image. (a) Original image; (b) compensation result by Wallis method; (c) local compensation result;(d) compensation result by original logarithmic transformation; (e) compensation result by the proposed method

Download full size

Index	Original valueof shadedarea	Non-shadedareatarget value	Compensationeffect whenV=0.5	Compensationeffect whenV=1	Compensationeffect whenV=2.5	Compensationeffect whenV=3
Meanbrightness	35.6813	61.5258	13.7326	26.5029	60.2695	70.3108
Meangradient	5.9157	10.7843	2.8213	4.3433	8.6223	9.9473

Table 1. Experimental data of logarithmic transformation

Index	Original value ofshaded area	Non-shaded areatarget value	Logarithmictransformation result	Localcompensation result	Joint compensationresult
Mean brightness	35.6813	61.3258	87.0096	43.8826	61.1334
Mean gradient	5.9157	10.5143	7.1328	11.9212	10.0059

Table 2. Image quality under each compensation method

Image	Index	Original valueof shadedarea	Non-shadedarea targetvalue	Compensationresult byWallismethod	Localcompensationresult	Compensationresult by originallogarithmictransformation	Compensationresult byproposedmethod
1	Meanbrightness	32.7592	57.3512	42.9142	42.8810	48.9746	57.3512
	Mean gradient	4.5116	7.2790	3.5462	8.6396	6.3676	8.9232
2	Meanbrightness	50.5660	86.8668	72.6850	65.1501	73.0049	86.2285
	Meangradient	7.6557	15.2537	6.1923	13.1376	10.0181	14.0568
3	Meanbrightness	29.1029	53.7794	37.5630	43.7125	49.7757	53.2503
	Meangradient	4.6264	9.1361	4.9575	8.8015	5.9548	8.4537
4	Meanbrightness	50.1019	98.8791	71.4916	75.1481	79.7161	95.5914
	Meangradient	6.0155	10.1674	7.6867	9.5654	7.9620	10.0674

Table 3. Evaluation of shadow compensation results

Image	Image size /(pixel×pixel )	Number of shaded pixels	Compensation time /s
Urban image 1	863×585	166541	1.13
Urban image 2	884×613	212022	1.62
Urban image 3	622×638	109438	1.01
Building image	257×266	9607	0.25

Table 4. Operation performance

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information