With the use of remote sensing imagery to model the natural phenomena such as disaster management mitigation, urban development etc., the need to assess the changes due to this phenomena is highly critical. The remote sensing data is complex in terms of the color and textural variations. Feature extraction is a crucial step in remote sensing image classification which directly affects the classification accuracy. The extraction of textural features on such images which are rich in spatial information is extremely important. There are many feature extraction techniques already available in the literature. But the main drawback of the existing techniques is the lack of obtaining the minute information embedded in the image. This is a critical stage in enhancing the change detection result. This paper presents a Dynamic Local Binary Pattern (DLBP) for textural feature extraction for satellite image change detection. The quantitative analysis of the proposed method is carried out by computing the kappa coefficient, false alarm and missed alarm.