Before CD can be attempted, the acquired images have to e registered to the site model. Two algorithms for image registration have been developed. When no information about the camera is available, an efficient constrained search mechanism is used for image-to-image registration. When an approximate camera model is available, as in RADIUS applications, a fast image-to-sitemodel registration algorithm is used which first projects the site model into the new image domain using the given approximate camera model, and then uses five control points to do camera resection and to obtain an accurate camera model. Using the image registration output, an image delineation algorithm has been developed which outputs regions of interest such as parking lots, roads or training grounds, depending on the underlying CD task. The approach has been illustrated by describing a site model supported image monitoring algorithm which uses contextual information, camera and illuminant models to detect and count vehicles or construction activities at the given area.
In related work, an energy function based approach to detecting rectangular shapes in images has been developed. The proposed edge-based approach involves extracting straight lines from an edge map of the image. Then a Markov Random Field (MRF) is built on these lines, i.e., a suitable neighborhood and an energy function are specified based on the relative orientation and spatial location of the lines. This energy function can be construed as a measure of the conditional probability of observing the lines given the rectangular shapes (the positions and number of which are unknown) in the image. Minimizing the energy function is equivalent to selecting the maximum likelihood estimate of the rectangular shapes in the image from the observed lines. simulated examples are presented to demonstrate the robustness of the proposed method. This approach, supplemented with some qualitative information about shadows and gradients, has been used to detect rectangular buildings in real aerial images. Due to poor quality of the real images, only partial shapes are extracted in some cases. A modified deformable contour (snakes) based approach can that be used for completion of the partial shapes.