A novel study on ant-based clustering for paddy rice image classification

摘要: Paddy rice is the major crop of food in Taiwan. There are three main contributions cultivation on Taiwan: regional eco-friendly environments, adjustment floods, and refreshing air. The estimation paddy area important since this information related to national policy, yearly yields calculation, post-disaster reimbursement. In past, a large amount human power time-consuming works performed through field exploration pattern revising. A more economic manner estimate desired. Accordingly, study aims design classifier by which evaluation remote sensing image can be rationally performed. study, ant-based clustering optimization (ACO) algorithm used present classification mapping. advantage that it effectively cluster data into groups unsupervised learning. ACO implementation organized following steps: (a) initialize number ants, (b) compute pheromone trail matrix (c) update matrix, (d) select best few ants have optimal object function value, (e) repeat iterations until termination criteria satisfied. meantime, self-organizing map (SOM) was as parallel for comparison. discernment built well thematic maps drawn. result shows has higher accuracy than SOM.