Shortwave Infrared-Based Phenology Index Method for Satellite Image Land Cover Classification
| Autor: | K. R. Sivabalan, E. Ramaraj |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
010504 meteorology & atmospheric sciences
Contextual image classification Pixel Computer science Feature extraction Multispectral image 0211 other engineering and technologies Confusion matrix 02 engineering and technology Land cover 01 natural sciences Statistical classification Satellite imagery 021101 geological & geomatics engineering 0105 earth and related environmental sciences Remote sensing |
| Zdroj: | Advances in Intelligent Systems and Computing ISBN: 9789811501838 SocProS (2) |
| DOI: | 10.1007/978-981-15-0184-5_75 |
| Popis: | Recent technology relay upon the satellites. Satellites are having their own technical property to its behavioral methods. Communication methodology and sensor techniques are very important to take earth imagery. Satellites take the images in various combinations of bands to produce the imagery with maximum details. The efficacy of the multispectral imagery is limitless. Multispectral imagery is having the details about a particular region using 11 band combinations. Grouping the pixels into the meaningful class is called as image classification. Satellite image classification is used to define the land surface segmentation and feature extraction. Weather forecasting, agriculture, air force department, and water department are the major departments which rely upon the satellite imagery. Remote sensing input for those real-time applications is endless. Satellite image classification falls with three categories as automatic, semi-automated, and hybrid methods. Medium resolution multi spectral imagery classification algorithms designed with Automatic and Manual classification method to reach the maximum accuracy. High and very high spectral imagery classification algorithms follow the hybrid- or object-based image analysis. Phonology index is designed to classify the multispectral satellite imagery based on the reflectance value of the passive sensor imagery. This paper provides a phenology-based method called shortwave infrared-based phenology index (SIPI) to classify the multispectral imagery with maximum accuracy. Confusion matrix and Kappa coefficient are the quality measures used to justify the classification efficiency. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|