Scribd
Upload
15 views3 pages

Thermal Anomaly Detection

anomaly detection

Uploaded by

Intoy Nako
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
15 views3 pages

Thermal Anomaly Detection

anomaly detection

Uploaded by

Intoy Nako
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

What is Anomaly Detection?

It looks very hard to explain the definition of “anomaly”. In our context, we don’t make any difference between anomaly and
outlier ( they are different actually, but we refer to same thing here).

But one example below gives you a first image introduction:

Actually, anomaly detection techniques attempt to find the objects that are “different” from the rest of the data objects in a
given data set. Usually, anomalies are generated from certain system mechanisms that are very different from the system
mechanism of the rest of data set.

XXXXXXXXXXXXOXXXX
XXXXXX
(see the example above?:)

What is anomaly detection ?


Formal definition:

Hawkins suggests that “An anomaly is an observation that deviates so much from other observations as to arouse suspicion
that is was generated by a different mechanism”. D. M. Hawkins, Identification of Outliers, 1st ed. Chapman and Hall,
London, 1980.
Chandola say that “Anomalies are patterns in data that do not conform to a well defined notion of normal behavior”. V.
Chandola, A. Banerjee, and V. Kumar, “Anomaly detection: A survey”, ACM Computing Surveys, vol. 41, no. 3, p.
ARTICLE 15, July 2009.

Knorr and Ng use very specific approach to define an outlier as “An object p in a data set D is a DB(pct, dmin)-outlier if at
least pct percent of the objects in D lie greater than distance dmin from p.”

In simple words: if something is very different from the other observations in the same data set with respect to the properties
that you have interests, then it is anomaly!

The problem of anomalies detection is of considerable importance, arising frequently in many different areas such as fraud
detection, cyber-intrusion detection, medical anomaly detection, image processing and textual anomaly detection.

 In financial area, banks spend millions of dollars on detecting credit card fraud and money laundering; using
anomaly detection techniques, they wish to identify abnormal usage patterns as soon as possible in order to prevent the
future loss.
 In cyber-intrusion field, companies use anomalies detection techniques to identify a hacker’s attacks by analyzing the
computer or website log files, and then attempt to identify abnormal user behaviors.
 In image processing, users can take advantage of anomaly detection algorithms to identify the image consisting of
different objects.

A thermal anomaly is defined as an object or feature that differs in temperature when compared,
spatially or temporally, with typical surface temperatures for other objects. That is, the surface
temperature may be irregular between one point and another (to form a spatial thermal anomaly)
or between one image and the next (to form a temporal thermal anomaly). It can also be defined
spectrally, in that (in both the spatial and temporal cases) it requires the presence of an object which
is significantly brighter or hotter than other objects present in the same spatial or temporal
dimension. Thus, to be resolvable the object has to be sufficiently different in terms of its spectral
radiance (i.e. be sufficiently bright or hot), as to be distinguishable, or resolvable, from other
surfaces in the spatial or temporal neighbourhood.

For the TIR

Because of the poor sensitivity of this waveband to sub-pixel hotspots, a small but hot sub-pixel
targets may not cause an anomaly in the TIR. However, the coincidence of this waveband with the
peak emission from surfaces at ambient temperatures means that it may be useful for detecting
subtle thermal anomalies associated with crater lakes or geothermal anomalies, especially in TM-
class data.

Can we resolve the hotspot?


Whether the anomaly is resolvable thus depends on (i) the waveband, (ii)the size and temperature
of the sub-pixel hot spot, and (iii) the degree to which the spectral radiance of the target pixel is
elevated above that of its surroundings. In effect, the background

Thermal Remote Sensing of Active Volcanoes: A User's Manual


By Andrew Harris

Cambridge University Press, Cambridge, New York, USA


Product Details
 Hardcover: 736 pages

 Publisher: Cambridge University Press (June 10, 2013)

 Language: English

 ISBN-10: 052185945X

 ISBN-13: 978-0521859455

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy