Abstract:The normalized difference vegetation index (NDVI) time-series database, derived from NOAA/AVHRR, SPOT/VEGETATION, TERRA or AQUA/MODIS, is increasingly being recognized as a valuable data source for extracting land cover and its change information at global, continental and large regional scale. However, existing approaches, such as principal component analysis (PCA) and change vector analysis (CVA) present considerable difficulties in taking full advantage the NDVI dataset for land cover change detection. Based on the assumptions that different land cover types have different NDVI temporal profiles and that the NDVI profile curve can be regarded as a spectrum in which an NDVI value for a certain date corresponds to on band value of this spectrum, we analyzed the existing change detection indexes and develop a new land cover change detection method based on Lance distance and a cross correlogram spectral matching (CCSM) technique. The new method was validated in the simulation experiments and a case study area of Beijing. From the results, we have demonstrated that the new method takes the shape and value features of NDVI profile curve into consideration. The relatively better performance of the new method can be attributed to two advantages: (1) the new method can discriminate long-term land cover changes form other changes by excluding false changes caused by vegetation phenology changes, climate events, atmospheric variability and sensor noise; (2) it is similarly sensitive to all kinds of land cover changes no matter where the changes have occurred. The better results compared with the CVA method suggest that the new method is effective and has potential for land cover change detection using an NDVI time-series dataset. Furthermore, it is worth noting that the method can not only be applied to NDVI datasets but also to other index datasets reflection surface conditions sampled at different time interval. It can also be applied to datasets for different satellites without the need to normalized sensor differences.

Study of land cover change detection based on texture features

Change Detection Method for High Resolution Remote Sensing Image in Association with Textural and Spectral Information

A Method of Land Use/land Cover Change Detection from Remote Sensing Image Based on Support Vector Machines

Application of Multi-Sensor Images for Detecting Land Cover Change and Analysis of Urban Expansion

Change Detection Using a Texture Feature Space Outlier Index from Mono-Temporal Remote Sensing Images and Vector Data.

Land cover change detection with a cross-correlogram spectral matching algorithm

Unsupervised Change Detection in HR Remote Sensing Imagery Based on Local Histogram Similarity and Progressive Otsu

REMOTE SENSING CHANGE DETECTION BY INCLUSION OF MULTITEMPORAL TEXTURE

Urban Land Use Change Detection Using Multisensor Satellite Images

Change Detection from Multi-Temporal Remote Sensing Images by Integrating Multiple Features

Sub-Pixel Change Detection For Urban Land-Cover Analysis Via Multi-Temporal Remote Sensing Images

Fast Detection of Changed Blocks in Land Use Map

A spectral gradient difference based approach for land cover change detection

Study on Land Cover Change Detection Method Based on Ndvi Time Series Batasets: Change Detection Indexes Design

THE APPLICATION OF MULTISCALE IMAGE TEXTURE TO THE DETECTION OF URBAN EXPANSION

A Method of Urban Change Detection Using High Spatial Resolution Remotely Sensed Data

Land-Use/Land-Cover Change Detection Using Improved Change-Vector Analysis

Detecting Spatio-Temporal and Typological Changes in Land Use from Landsat Image Time Series

Land Cover Change Detection by Integrating Object-Based Data Blending Model of Landsat and MODIS

MIDDLE RESOLUTION REMOTE SENSING IMAGE CHANGE DETECTION BASED ON VECTOR ANALYSIS OF MIDLINE CHANGE

Change Detection Method for High Resolution Remote Sensing Images Using Deep Learning