Application Of Knowledge-Based Image Classification And Ca -Markov Chain Prediction Model For Landuse / Landcover Change Analysis Of Onitsha And Environs, Anambra State

Knowledge-based classification have proved to be effective for complex object recognition and for image analysis Therefore, this study was aimed on carried out land use/ land cover change detection and prediction using knowledge-based classification and Cellular Automata (CA)_MARKOV model of Onitsha. Landsat images of 2008, 2013 and 2018 covering the study area was acquired and used to carryout LULC classification and change detection. The change detection was carried out using the matrix union overlay in ERDAS 2014. Result shows that build-up area increased from 39.4% to 43% and from 43% to 45.9% between year 2008 to 2013 and 2013 to 2018 respectively. The water body increased in year 2013 from 6.5% to 7.2%, then decreased in 2018 to 6.8%. While the vegetation keeps decreasing all through the year from 42% in year 2008 to 36.1% in year 2018. CA-MARKOV model was then used to predict the landuse/land cover changes in the study area to year 2025. The result shows that built-up area will increase from 45.9% to 48% and also, decreases in vegetation, open space, sand dunes and water body. The study Recommend that Knowledge based classification should be used as it gives a better understanding and classification for land use / land cover types.