This document discusses supervised and unsupervised image classification techniques.
Supervised classification involves (1) a training stage where sample pixels are used to develop land cover class descriptions, (2) a classification stage where each pixel is assigned a land cover class based on spectral similarity to the training data, and (3) an accuracy assessment comparing the classification to ground truth data. Unsupervised classification clusters pixels without prior class definitions, requiring the analyst to interpret the clusters. Both approaches aim to categorize remotely sensed images but differ in how the categories are defined.
This document discusses supervised and unsupervised image classification techniques.
Supervised classification involves (1) a training stage where sample pixels are used to develop land cover class descriptions, (2) a classification stage where each pixel is assigned a land cover class based on spectral similarity to the training data, and (3) an accuracy assessment comparing the classification to ground truth data. Unsupervised classification clusters pixels without prior class definitions, requiring the analyst to interpret the clusters. Both approaches aim to categorize remotely sensed images but differ in how the categories are defined.
This document discusses supervised and unsupervised image classification techniques.
Supervised classification involves (1) a training stage where sample pixels are used to develop land cover class descriptions, (2) a classification stage where each pixel is assigned a land cover class based on spectral similarity to the training data, and (3) an accuracy assessment comparing the classification to ground truth data. Unsupervised classification clusters pixels without prior class definitions, requiring the analyst to interpret the clusters. Both approaches aim to categorize remotely sensed images but differ in how the categories are defined.
This document discusses supervised and unsupervised image classification techniques.
Supervised classification involves (1) a training stage where sample pixels are used to develop land cover class descriptions, (2) a classification stage where each pixel is assigned a land cover class based on spectral similarity to the training data, and (3) an accuracy assessment comparing the classification to ground truth data. Unsupervised classification clusters pixels without prior class definitions, requiring the analyst to interpret the clusters. Both approaches aim to categorize remotely sensed images but differ in how the categories are defined.
Download as PPTX, PDF, TXT or read online from Scribd
Download as pptx, pdf, or txt
You are on page 1of 15
Image Classification: Supervised Classification
• Image classification refers to the computer-assisted
interpretation of remotely sensed images. • The objective of image classification procedures is to automatically categorize all pixels in an image into land cover classes or themes. • For categorization, the spectral pattern present within the data for each pixel is used as the numerical basis. • The term pattern refers to the set of radiance measurements obtained in the various wavelength bands for each pixel. The family of classification procedures can be categorized into • Spectral pattern recognition • Spatial pattern recognition • Temporal pattern recognition 1. Spectral pattern recognition:- • This procedure utilizes the pixel by pixel spectral information as the basis for automated land cover classification.
2. Spatial pattern recognition:-
• This procedure involves the categorization of image pixels on the basis of their spatial relationship with pixels surrounding them and as per the tonal variation. 3. Temporal pattern recognition:- • This procedure uses time as an aid in feature identification. Data is analyzed from imagery recorded on different dates. • This is particularly relevant in the case of crop surveys as their imagery will go through changes during the growing season. • These procedures can be combined when the need arises. Depending on the nature of the data being analyzed, the computational resources available and the intended application of the classified data, the approach or the procedure can be arrived at. 1. Training stage in Supervised Classification
• The analyst identifies representative training areas
and develops numerical descriptions of the spectral signatures of each land cover type of interest in the scene. This is also called signature analysis. • It also requires substantial reference data and a thorough knowledge of the geographic area to which the data apply. • The quality of the training process determines the success of the classification stage and thereby the value of the information generated from the entire procedure Technicality 2. Selection of Training Data
• Accurate selection of training data is crucial
for accurate supervised classification.
• There are many approaches to signature
collection and analysis, but all rely to a certain degree on the experience of the analyst. 3. The output stage:- • The typical forms of output products are thematic maps, tables and digital data files, which become input data for GIS. • The figure given below shows the flow of operations to be performed Unsupervised classification
• This procedure examines the data and breaks it into the
most prevalent natural spectral groupings, or clusters, present in the data. • The analyst then identifies these clusters as land cover classes through a combination of familiarity with the region and ground truth visits. The logic by which unsupervised classification works is known as cluster analysis. • In contrast to supervised classification, where the system needs to be told about the character (i.e., signature) of the information classes we are looking for, unsupervised classification requires no advance information about the classes of interest. Continue………
• It is important to recognize, however, that the clusters
unsupervised classification produces are not information classes, but spectral classes (i.e., they group together features (pixels) with similar reflectance patterns). • It is thus usually the case that the analyst needs to reclassify spectral classes into information classes. For example, the system might identify classes for Road and Residential which the analyst might later group together, creating an information class called Built-up. • The quality of the classification will depend upon the analyst’s understanding of the concepts behind the classifiers available and knowledge about the land cover types under analysis. Classification accuracy assessment
• Classification accuracy assessment is a general term
for comparing the classification to geographical data that are assumed to be true to determine the accuracy of the classification process. • Usually, the assumed true data are derived from ground truth. It is usually not practical to ground truth or otherwise test every pixel of a classified image. • Therefore a set of reference pixels is usually used. Reference pixels are points on the classified image for which actual data are (will be) known. The reference pixels are randomly selected. Continue………..
• Once a classification exercise has been carried
out there is a need to determine the degree of error in the end-product. These errors could be thought of as being due to incorrect labeling of the pixels. • The basic idea is to compare the predicted classification (supervised or unsupervised) of each pixel with the actual classification as discovered by ground truth. The Confusion Matrix (Error Matrix)
• The most commonly-used method of representing the
degree of accuracy of a classification is to build a confusion (or error) matrix.
• The analyst selects a sample of pixels and then visits the
sites (or vice-versa), and builds a confusion matrix. This is used to determine the nature and frequency of errors.
• Overall map accuracy = total on diagonal / grand total
I would like to say thanks for paying attention in all my lectures and Labs & Hope you like the way I taught, If don’t….. then forget it……ha ha haaaa & The main thing is Lord Bless you, study more and more & Last but not the least Best of luck for the finally examination