International Research Journal of Engineering and Technology (IRJET) Volume: 04 Issue: 05 | May-2017
www.irjet.net
e-ISSN: 2395 -0056 p-ISSN: 2395-0072
SHADOW DETECTION AND REMOVAL OF PANCHROMATIC SATELLITE IMAGES Amrutha Vishnupriya S1, Vicky Nair2 1P G
scholar, Dept. of Computer Science And Engineering, Musaliar College of Engineering and Technology, Pathanamthitta
2 Associate
professor, Dept. of Computer Science And Engineering, Musaliar College of Engineering and Technology, Pathanamthitta
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract -
the high-resolution satellite images contain shadows as undesired information which causes partial or total loss of information. In such a circumstance, the objects in the shadow regions are difficult to be extracted for further processing or applications. So, in order to restore obscured objects, shadow detection and shadow removal is an essential preprocessing step for panchromatic highresolution satellite images. Many effective algorithms of shadow removal have been proposed for natural images or remote sensing multispectral images. However, there is a great lack of shadow removal method for panchromatic imagery, while these panchromatic images usually contain high resolution data to be useful for various applications. For the purpose of the information recovery of obscured objects, the characteristics of shadows and objects in the panchromatic images of urban areas should be analyzed and remove the shadows to obtain shadow-free images. Many current researches and studies indicate that shadow detection is the indispensable step in the complete processing chain of shadow removal.
The existence of shadows in panchromatic satellite images can occlude some objects in images to cause the reduction or loss of their information, particularly in urban scenes. Shadow removal is a significant processing procedure for recovering the occluded information of objects, which is useful for many object oriented applications. This system is a novel framework of shadow detection and removal for panchromatic satellite images to restore the obscured object information. In shadow detection module , hard shadow detection and soft shadow detection method are used by the combined application of a multi - thresh holding method and image matting technique. Soft detection results can show both umbra areas and penumbra areas to describe the shadow distribution precisely and this result is used for removing the detected shadow in two levels. In the initial correction of shadow removal, the linear correlation is used to enhance shadow areas roughly in global. In the second level refined process, the characteristics of intensity of objects and shadows are noted and patch intensity difference is calculated and it is used to remove the shadow portion. The linear radiometric correction and intensity adjustment are used to control the brightness. Finally the shadow free image is smoothened by applying Gaussian filter.
Key Words: Gaussian filter, Image matting, linear correlation method, Panchromatic satellite images, Penumbra, Soft shadow detection, Umbra.
In the proposed system, build a novel shadow detection and removal method has been developed for the panchromatic satellite images.Multilevel image thresholding method and image matting technique are both used to auobtain soft shadow detection results. In the initial step of shadow removal, the linear correlation method is used to enhance shadow areas roughly in global. In the second level refined process, the characteristics of intensity of objects and shadows are noted and patch intensity difference is calculated and it is used to remove the shadow portion. The linear radiometric correction and intensity adjustment are used to control the brightness. Finally the shadow free image
1. INTRODUCTION The high resolution Satellite imaging is recommended and necessary for the observation of space, earth and other planets has ability to obtain very highresolution images. These high-resolution images exhibit more detail information to increase the object-oriented applications, like building detection and to spot out populated and non-populated areas. Unfortunately, most of
Š 2017, IRJET
|
Impact Factor value: 5.181
|
ISO 9001:2008 Certified Journal
|
Page 496