IDENTIFICATION OF PADDY FIELD FROM VERY HIGH RESOLUTION IMAGE USING OBJECT BASED IMAGE ANALYSIS METHOD. (A CASE STUDY IN RANCAEKEK, BANDUNG, WEST JAVA, INDONESIA)

摘要

The calculation of existing land of food is very important to estimate the national food security. In this case remote sensing can be used as better alternative method in the paddy field identification. The study focuses on paddy field identification using very high resolution imagery, Quickbird imagery acquired in 2008, and applying OBIA (Object Based Image Analysis) as a method for analysis of paddy field in Rancaekek, Bandung, Indonesia. To identify an object, this method not only analyse the value of a pixel but also to consider the neighbour pixels. By the application of OBIA method in identifying wetland that it considers many parameters such as land form (shape), soil characteristics (spectral), land texture, and the relationship between wetland and other objects around them. The study found that OBIA method can be used to identify fields of raw land, generating a clear delineation of fields and also classified wetland according to the characteristics of paddy growth phase (water phase, vegetative, generative, and harvest phases) with accurate. Finally, the calculation of paddy field can be estimated from the results of image classification.