Abstract: Digital surface Models (DSMs) contain information about topographic surfaces and also about other objects higher than the surrounding topographic surface such as buildings, trees etc. therefore, they are ideal for analyzing urban landscape. In this research DSMs of Bhopal city are made from Cartosat 1 stereo data. Three DSMs are made using three different techniques such as adaptive ATE, traditional ATE available through Leica Photogrammetric Suite (LPS) and automated terrain extraction available through Orthoengine. The paper compares the steps of DSM extraction and concluded that process of DSM generation comprised of image orientation and image matching. The output DSMs are compared in context of identifying urban features. It was found that DSM generated through traditional ATE and Orthoengine are more reliable compared to DSM generated through adaptive ATE. This is because; object filter within adaptive ATE modifies the elevation value resulting error in object heights.