Effect of Different Spatial Resolutions of Multi-temporal Satellite Images Change Detection Application

Abstract

Remote sensing (RS) satellites may provide much information about the land cover at different resolutions that have been utilized in many military and civil purposes. Multi-temporal images Change detection (CD) is one of the RS applications. Images Co-registration is an important step before the change detection process, and although they are geo-referenced to each other they may not have the same spatial resolution. So, prior to the multi-temporal analysis, images should be similar in pixel size. Using optical sensors with different characteristics and resolutions to obtain the same geographical area may cause effects in change detection results. In this paper; Image difference pixel-based change detection technique is proposed, and the effect of the image pixel size on change detection is studied. Two Images of the same area are taken from two different sensors; Worldview2 and Quickbird2 images with “2”m and “2.4”m pixel size are used respectively. The change detection results show that; for “2” m pixel size, the change detection is 0.15 % With regard to “1” m, the change detection is 0.06 % With regard to “2.4” m, and 0.05 % with regard to “4” m pixel size. On the other hand, the change detection results show that, for “2.4” m pixel size, the change detection is 0.09 % With regard to “1” m and 0.015 % With regard to “4” m pixel size.