Remote Sensing Digital Image Analysis

Abstract

In the time since the second edition of this text was produced two significant trends have been apparent. First, access to image processing technology has continued to improve significantly with most students and practitioners now having readily available inexpensive workstations and powerful software for analysing and manipulating image data.The second change has been the dramatic increase in the numbers of satellite, aircraft and sensor programs. Perhaps most significant is the widespread availability of hyper spectral data and the special challenges presented by that data type for information extraction.Accordingly, this third edition has been written to reflect those trends while, at the same time, preserving the important elements of image processing and analysis algorithms of significance in remote sensing applications.The major changes between the previous edition and this are an update of Chapter 1, and the introduction ofa new Chapter 13 dealing with methods for analysing hyperspectral data sets.Chapter 12 has also been significantly altered to provide a focus on the interpretation of data sets that are mixed and could contain, for example, different types of imagery along with other spatial data types found in a geographic information system. The previous knowledge-based material has been retained but material originally covered in Chapter 8 dealing with multi-source data analysis has been combined with knowledge-based methods to create this chapter on data fusion.The authors wish to express their appreciation to their colleagues for the assistance they have received in preparing this new edition. In particular David Langrebe of Purdue University remains a great supporter and both authors have had the benefit of working with Dave over the years, including periods spent at Purdue.The authors are also enormously grateful to their families for their understanding and support in making the completion of this third edition possible. Canberra, Australia, August 1998John A. Richards Xiuping Jia VII support and assistance. Ashwin Srinivasan, now with the Turing Institute in Glasgow, as a very gifted graduate student, developed much of the material on which Chapter 12 is based and, during his time as a student, helped the author understand the mechanisms of knowledge processing. He also kindly read and offered comments on that chapter. The author's colleague Don Fraser similarly read and provided comments on the material on neural networks. Philip Swain and David Landgrebe of Purdue University continue to support the author in many ways: through their feedback on the fIrst edition, their interaction on image processing, and Dave's provision of MultiSpec, the Macintosh computer version of the LARSYS software package. Finally, the author again expresses gratitude to his family for their constant support, without which the energy and enthusiasm needed to complete this edition might not have been found.

Comparison of Pixel-Based and Object-Based Classification Using Parameters and Non-Parameters Approach For the Pattern Consistency of Multi Scale Landcover

“…"Digital classification is quantitative analysis of computer interpretation to identify pixel-based upon their numerical properties and owing to its ability for counting pixel for area estimates" (Richards and Xiuping, 2006). Following classification methods implemented to such data, and the result would be quantify statistically and spatially to assess pattern consistency.…”

“…MLC is the most common supervised classification, this classifier using parametric approach which the spectral statistic data should be distributed normally, and also the decision rule in MLC to estimate from training data determined by Bayes" theorem (Richards and Xiuping, 2006). MLC use "mean value and covariances to compute the probability of individual pixels belonging to specific class for the variation in spectral data" (Ghosh and Joshi, 2014).…”

ProSeg: a database of local structures of protein segments

“…These 370 proteins were selected as a set of non-redundant representative proteins from the Culled Protein Data Bank (PDB) (version: Dec. 13,2001; resolution \1.6 Å ; R-factor \ 0.2; sequence identity \25%) [8]. The divided segments were classified into a number of clusters by a single-pass clustering (SP) method [9] or a 3D mesh gridding (3DMesh) method. Structural dissimilarity (D or D issim ) between the segments is defined on the basis of backbone dihedral angles.…”

Görüntü ve lidar verisinden bina tespitinde farkli yöntemler

“…En büyük benzerlik sınıflandırması eğitimli sınıflandırma çeşitlerinden en sık kullanılanıdır [13]. Ek bilgilerin de mevcut olduğu zaman, örneğin Sayısal yükseklik/arazi modeli veya kadastral plan gibi, sınıflandırma sonucu bir takım kurallar tanımlanarak, birbirine karışmış sınıfların ayrıştırılması veya birleşmesi gereken sınıfların birleştirilmesiyle iyileştirilebilmektedir.…”