Abstract: Now a days, there has been an extensive increase in the availability of high-resolution commercial satellite imagery, enabling a variety of new remote-sensing applications. Object detection is an essential part in object recognition for spatial information systems. Spatial object are formed by psychological and mathematical two important factors. Geographical object based image analysis is widely used in obtaining the information from remotely sensed images. The expert’s knowledge is used to incorporate the extracted information. In this paper, we present a novel approach for detecting objects in various types of satellite images such as: Visible Satellite Images, Infrared Satellite Images and Water Vapor Satellite Images and analyze the natural importance of the objects. Our approach requires a learning of the specific structures wish to detect using a prior knowledge about them. For this, a set of different images for the segmented images of the three types of satellite images is provided to the human subjects, such that they can improve their understanding using their perception, cognition and decision for the spatial object recognition. The result obtained for two stage of image processing is collected and a relationship to psychological and mathematical basis is made. The results show that certain association pertinent scale to the human perception is helping more to recognize the spatial objects. The experiment is carried out in MATLAB software where the results shown that our approach is better eligible for accurate object detection and recognition on different kinds of satellite images.