Automatic building extraction remains an open research topic in digital photogrammetry and remote sensing. While many algorithms have been proposed for building extraction, none of them solve the problem completely. This is even a greater challenge in urban areas, due to high-object density and scene complexity. Standard approaches do not achieve satisfactory performance, especially with high-resolution satellite images. This paper presents a novel framework for reliable and accurate building extraction from high-resolution panchromatic images. Proposed framework exploits the domain knowledge (spatial and spectral properties) about the nature of objects in the scene, their optical interactions and their impact on the resulting image. The steps in the approach consist of 1) directional morphological enhancement; 2) multiseed-based clustering technique using internal gray variance (IGV); 3) shadow detection; 4) false alarm reduction using positional information of both building edge and shadow; and 5) adaptive threshold based segmentation technique. We have evaluated the algorithm using a variety of images from IKONOS and QuickBird satellites. The results demonstrate that the proposed algorithm is both accurate and efficient.
Object classifier often operates by making decisions based on the values of several shape properties measured from an image of the object. The paper introduces a unique definition of measure for 2-D geometrical object shape detection. Using this definition different object shapes can be identified on the basis of their degree of fitness parameter. Basically, we have fitted a 2-D polygon/curve on the object as a best fitted polygon/curve and computed the parameter degree of fitness which is the ratio of the matching area and non-matching area due to the fitted polygon/curve and the object both. The results show the effectiveness of the proposed measure.
Automatic target detection like oil tank from satellite based remote sensing imagery is one of the important domains in many civilian and military applications. This could be used for disaster monitoring, oil leakage, etc. We present an automatic approach for detection of circular shaped bright oil tanks with high accuracy. The image is first enhanced to emphasize the bright objects using a morphological approach. Then, the enhanced image is segmented using split-and-merge segmentation technique. Here, we introduce a knowledge base strategy based on the region removal technique and spatial relationship operation for detection of possible oil tanks from the segmented image using minimal spanning tree. Lastly, we introduce a supervised classifier, for identification of oil tanks, based on the knowledge database of large amount data of oil tanks. The uniqueness of the proposed technique is that it is useful for detection bright oil tanks from high as well as low resolution images, but the technique is always better for high-resolution imagery. We have systematically evaluated the algorithm on different satellite images like IRS -1C, IKONOS, QuickBird, and CARTOSAT -2A. The proposed technique is detected bright structures but unable to detect the dark structure. If the oil tank structures are bright relative to the background illumination in the image then the detection accuracy by the proposed technique for the high resolution image is more than 95 per cent.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.