VISUAL AND STATISTICAL QUALITY ASSESSMENT AND IMPROVEMENT OF 
REMOTELY SENSED IMAGES 
S.Mohammad Shahrokhy 
Iran's Space Agency- Tehran — Iran 
s m sh(ghotmail.com 
KEYWORDS: Atmospheric, Radiometric, Geometric, QualityDiagnosis, Assessment, Improvement Elimination 
ABSTRACT: 
Remotely sensed images are interpreted pixel by pixel, using spectral vector analysis methods. Most kind of noise and perturbation in 
pixel value or position cause misinterpretation. In this paper most common Radiometric, Atmospheric and Geometric defects of 
remotely sensed images are investigated along with the diagnosis and elimination methods on some high and medium resolution satellite 
images. Quality assessment isperformed in both visual and statistical manner and also qualitymprovement is fulfilled in both Manual 
and Automatic ways. Many technical methodsare used such as histogram transformation, mean, variance and median calculation of lines 
and bands, spatial filtering, template matchingyectifications using GCPs and brghtness temperature and reflectance checking. Visual 
diagnosis of defects isoften more precise but not appropriate for automatic procedures. Manual elimination of the defects is also more 
accurate however time consuming and user dependent. 
1. INTRODUCTION 
  
  
  
; : ; bass ; uality Defect Visual Diagnosis Statistical 
Reliable interpretation and results necessitateinput data Quality Q y > Diagnosis 
Assessment (QA) and sometimes Quality Improvement (QD). Striping Different overall Significandy 
On the other hand, in automatic procedures, image Quality 
should be checked to accept or reject the input or sometimes 
improve itto be able to cope with the expectedduty. 
Remote sensing image Quality generally has three aspects 
Radiometric Quality, Atmospheric Quality and Geometric 
Quality. Radiometric Quality is affected by sensor 
Le eo . S c € right € SITIC 
characteristics and detector responses. Striping Drop lines, Noise Dark amd bright Radiometric 
. n EM eL 3 es points at the anomalies 
Noise and Band missing are of this sort. AtmosphericQuality is 
; ; ] ; background 
dependent on the circumstances at the imaging time. Cloud T = 7 
= as ; n° he Band Missing Lack of data in a Zero variance of a 
cover and Haze are of this type. Geometric Quality is either band Band 
at an 
dependent on sensor characteristics and also satellite sitation 
such as attitude, position, velocity and perturbations. Earth} 
surface relief is another important factoraffecting Geometric 
Quality of the image. Band to band Misregistration and image 
to map Misregistration areof geometric Quality elements (QE) 
[t is essential to note that each sensor has special Quality 
Assessment and Quality Improvement methods, thresholds and 
coefficients So images of each sensor must be processed 
separately. In this research, TERRA-MODIS, NOAA-AVHRR, 
IRS-PAN and IRS-LISS III images are investigated. 
Many works have been done on image Quality control 
(Barrett 1990, Nill & Bouzas 1992, Eskicioglu & Fisher 1995, 
Barrett 1995, Westen et al 1995, Taylor 1998, 
Avicibasand Sankur 2000) and generally each company 
provides a compbte report of its sensor imags and products 
Quality e.g. EOS (Chu et al 2000 , Vermote et al 1997). 
2. QE AND DEFECT DIAGNOSIS 
2.1 Radiometric Quality Assessment 
Radiometric Quality elements and recognition methods are 
briefly listedin Table 1. 
104 
brightness of 
adjacent lines 
different variance 
and mean of 
adjacent lines 
  
Drop Line 
Null scan line 
Zero variance of a 
line 
  
  
  
  
  
  
  
Table 1. Radiometric Qualitydefects and diagnosis methods 
Striping is caused by different response of elements of a 
detector array to same amount of inoming EM energy. This 
phenomenon causes heterogeneity in overall brightness of 
adjacent lines figure 1). 
  
b 
hd 
Figure I. Image No.l (MODIS) with stripes 
  
in à