A Simple Shadow Area Processing Method

Haiqing Wang

China Aero Geophysical Survey and Remote Sensing Center for Land and Resources,Beijing 100083,China.

Email: whq0705@126.com

Keywords: Remote sensing, shadow area, ratio analysis, image processing

Abstract: Shadow is an important factor that restricts remote sensing information extraction. How to use simple and

effective image processing method to display the remote sensing information of shadow area has been a

difficult problem in the field of remote sensing. In this paper, a simple ratio analysis method is applied to

study the shadow area remote sensing image processing, which shows the remote sensing information

hidden in the shadow area better. The method lays a good foundation for further remote sensing information

extraction. This method is simple and effective, not only can solve the problem, but also easy to operate,

even the non-remote sensing image processing professionals can also be used flexibly.

1 INTRODUCTION

Shadow is an important factor that restricts remote

sensing information extraction. How to use simple

and effective image processing method to display

the remote sensing information of shadow area has

been a difficult problem in the field of remote

sensing. There are many research findings about

shadow processing, in recent years. A shadow

processing method based on normalized RGB colour

model was proposed by Yang and Zhao (Yang and

Zhao, 2007).A shadow compensation method based

on linear stretching, smoothing and principal

component was proposed by Wang and Wang

(Wang and Wang, 2010). By improving the Wallis

filtering shadow compensation strategy, the ground

information in the shadow area was highlighted by

Gao et al. (Gao et al., 2012). The shadow vegetation

index SVI was constructed to discuss the problem of

image shadow removal by Xu et al. (Xu et al.,

2013).Combining the wave band regression model

and shadow vegetation index SVI can be effective,

according to Liu et al. (Liu et al., 2013). Gao et al.

(Gao et al., 2014) believe that in order to

compensate the model as the foundation, through the

mean brightness shadow and non-shadow region

statistics and variance, it is possible to use the

method of feature extraction and matching of

automatic acquisition of model parameters,

automatic compensation and shadow comprehensive

regional overall level of compensation and

compensation for the two level local window. Deng

et al. (Deng et al., 2015) explored the use of blue

light suppression algorithm and statistical

information of shadow homogeneity to compensate

for H, I and S components, respectively and

converted the results to RGB colour space to

achieve shadow compensation. Based on ArcGIS

Engine platform, Matlab and GDAL development

tools, Yang et al. (Yang et al., 2015) integrated

shadow detection and compensation systems

designed according to the shadow detection and

compensation algorithm of high resolution remote

sensing images. Shadow removal of remote sensing

images based on inhomogeneous regularized

texture-preserving was proposed by Fang et al.

(Fang et al., 2015). The shadow removal model of

traditional HSV space by integrating one step

information, based on it, a shadow removal

algorithm of moving objects based on reflectance

ratio invariants, was proposed by Zhang and Yang

(Zhang and Yang, 2016). Improvement of image

shadow tracking and elimination algorithm based on

texture loss least constraint was proposed by Yan et

al. (Yan et al., 2016). Methods of pattern recognition

and image enhancement are used to discuss the

problem of shadow removal by Zhao et al. (Zhao et

al., 2016). Methods used shadow extraction,

envelope removal, similar pixel search and shadow

brightness reconstruction to explore the shadow

Wang, H.

A Simple Shadow Area Processing Method.

In Proceedings of the International Workshop on Environment and Geoscience (IWEG 2018), pages 507-510

ISBN: 978-989-758-342-1

507

information reconstruction experiment of Landsat 8

OLI image in hilly area of southwest China were

proposed by Zhang et al. (Zhang et al., 2017).

Shadow removal algorithm based on improved

Gaussian mixture model and texture was proposed

by Wang and Zhang (Wang and Zhang., 2017). A

method of target detection and shadow removal

based on the combination of improved adaptive

hybrid Gaussian model and colour space was

proposed by Wang and Tong (Wang and Tong,

2013). These shadow removal methods are more

suitable for remote sensing researchers, and master

software professionals, however for remote sensing

interpretation and information extraction of

personnel engaged in the application of remote

sensing, these software are often not good, and need

a lot of energy to carry on the software

programming, it is also difficult to extract remote

sensing information of a shadow area. If a simple,

effective, easy to understand and easy to operate

shadow area processing method could be found; it

will solve the problem of remote sensing shadow for

interpreters and information extraction personnel.

2 AN OVERVIEW OF THE

STUDY AREA

The study area is located in the middle of Qinghai

Province, in the southeastern side of the Qaidam

Basin. It is a part of Kunlun mountain, Burhan

Budai Mountains in China, the highest peak of

which is 5000 meters above sea level. The study

area is located at an elevation of 4000 meters. Under

the action of plate movement, the ground surface

continues to uplift, the erosion is sharp and the

terrain is rugged where it is located at the southern

margin barrier of the Qaidam Basin. The remote

sensing image of the mountain slope often has

shadow, causing a lot of trouble to the remote

sensing information extraction and interpretation

work, need a simple and effective treatment method

of shadow zone.

3 REMOTE SENSING DATA AND

PRE-PROCESSING

In order to improve remote sensing image of shadow

area, the GeoEye-1 and Worldview-2 remote

sensing data of the study area were obtained (Table

1).

In order to ensure that the research work will be

carried out smoothly, first of all remote sensing data

quality was checked for the vegetation cover, the

amount of snow and ice, cloud cover, distortion, and

noise. Inspection shows that the remote sensing data

is characterized by rare cloud and snow, low

vegetation cover, no distortion, no noise (Table 2).

Table 1: List of remote sensing data and their characteristics.

Number

Data Type

Band Name

Band Number

Resolution(m)

Spectral Range(nm)

GeoEye-1

Pan

0.5

450-800

Blue

450-510

Green

510-580

Red

655-690

Near Infrared

780-920

Worldview-2

Pan

0.5

450-800

Coastal

400-450

Blue

450-510

Green

510-580

Yellow

585-625

Red

630-690

Red Edge

705-745

Near Infrared 1

770-895

Near Infrared 2

860-1040

Table 2: List of remote sensing data inspection

Number

Data Type

Vegetation

Snow and Ice

Cloud

Distortion

Noise

Strip

GeoEye-1

Very Sparse

<5%

Not Obvious

Worldview-2

Very Sparse

<5%

Not Obvious

IWEG 2018 - International Workshop on Environment and Geoscience

508

Figure 1: Image contrast of GeoEye-1 before and after data ratio operation (Left: B1; Middle: B3; Right: B1/B3).

Figure 2: Image contrast of Worldview-2 before and after data ratio operation (Left: B3; Middle: B8; Right:

B3/B8).

Remote sensing data pre-treatment includes raw

data normalization, image rectification, band

registration, image mosaic, data fusion, removal of

interference and tailoring of black edges. The

processed remote sensing data are more suitable for

subsequent research.

4 IMAGE PROCESSING IN

SHADOW AREA

In order to display remote sensing information

hidden in the shadow area, a variety of remote

sensing image processing methods had been tried,

and it was found that the method of using band ratio

operation is simple and effective.

Using GeoEye-1 remote sensing data to conduct

shadow area remote sensing image processing

method, it was found that after executing B1/B3

band ratio operation, the hidden information in the

shadow area can be displayed, which is convenient

for remote sensing interpretation (Figure 1).

Using Worldview-2 remote sensing data to

conduct shadow area remote sensing image

processing method, it is found that after B3/B8 band

ratio operation, the hidden information in the

shadow area can be displayed, which is convenient

for remote sensing interpretation (Figure 2).

5 CONCLUSIONS

It is concluded that by using GeoEye-1 and

Worldview-2 remote sensing data, B1 to B3 and B3

to B8 band ratios respectively can make the hidden

information in the shadow area appear, which makes

the interpretation of the remote sensing images

easier.

This method is simple and effective, not only to

solve the shadow problem, but also easy to operate,

even for the non-remote sensing image processing

professionals.

A Simple Shadow Area Processing Method

509

ACKNOWLEDGMENTS

China Geological Survey Project: national mineral

resources development environment remote sensing

monitoring (DD20160075, 121201003000172705),

national mine new restoration and treatment

monitoring in 2017 (DD20189805,

121201003000172718) support.

REFERENCES

Deng L, Deng M J and Zhang L S 2015 Optimization of

shadow detection and compensation method for high-

resolution remote sensing images J Remote Sensing

Technology and Application 30(2) 277-284

Fang Y M, Li B L and Wu Y J 2015 Shadow removal of

remote sensing images based on inhomogeneous

regularized texture-preserving J Journal of Computer

Applications 35(S2) 238-241

Gao X J, Wan Y C and Yang Y W 2014 Automatic

shadow detection and automatic compensation in high

resolution remote sensing images J Acta Automatica

Sinica 40(8) 1709-1720

Gao X J, Wan Y C and Zheng S Y 2012 Automatic

shadow detection and compensation of aerial remote

sensing image J Geomatics and Information Science

of Wuhan University 37(11) 1299-1302

Liu J, Xu Z H and Yu K Y 2013 Shadow detection and

removal method for remote sensing image in

mountainous and hilly area J Transactions of the

Chinese Society for Agricultural Machinery 44(10)

238-241, 237

Wang H R and Tong W 2017 Target detection and shadow

removal method based on the improved gaussian

model J Journal of Jiangsu University of Science and

Technology (Natural Science Edition) 31(2) 172-177

Wang Y and Wang S G 2010 Detection and compensation

of shadows in high resolution remote sensing images

using PCA J. Journal of Applied Sciences 28(2) 136-

141

Wang Y N and Zhang H 2017 Shadow removal algorithm

based on improved gaussian mixture model and

texture J Computer Engineering and Design 38(7)

1879-1884

Xu Z H, Liu J and Yu K Y 2013 Construction of

vegetation shadow index (SVI) and application effects

in four remote sensing image J Spectroscopy and

Spectral Analysis 33(12) 3359-3365

Yan F, Zhang J and Wu S D 2016 Improvement of image

shadow tracking and elimination algorithm based on

texture loss least constraint J Modern Electronics

Technique 39(27) 104-108

Yang J and Zhao Z M 2007 Shadow processing method

based on normalized RGB color model J. Opto-

Electronic Engineering 34(12) 92-96

Yang X W, Yang S W and Zhang L M 2015 Design and

development of the shadow detection and

compensation system for high-resolution remote

sensing images J Remote Sensing for Land and

Resources 27(3) 177-181

Zhang P and Yang Y X 2016 Multiple information fusion

of video moving target shadow removal algorithm J

Video Engineering 40(2) 59-64

Zhang T, Liao H P and Cui L L 2017 Mountainous

shadow information restoration based on the

continuum removed J Journal of Remote Sensing 21(4)

604-613

Zhao Y F, Ren H E and Hu J F 2016 Genetic algorithm

and homomorphic filter for log face image process J

Journal of Northeast Forestry University 44(8) 92-96

IWEG 2018 - International Workshop on Environment and Geoscience

510