1 August 1990 Artificial neural networks for photogrammetric target processing
Author Affiliations +
Proceedings Volume 1395, Close-Range Photogrammetry Meets Machine Vision; 13952U (1990) https://doi.org/10.1117/12.2294345
Event: Close-Range Photogrammetry Meets Machine Vision, 1990, Zurich, Switzerland
The ultimate aim of this work is to produce a control system for two dimensional measuring machines which will be capable of automatically seeking and finding 'target' images on photogram- metric plates, giving accurate estimates of the co-ordinates of the centres of the target and some objec- tive measure of the uncertainty of these estimates. It is well known that the target locating process is labour intensive, generally requiring days of work for operators to identify several thousand targets contained in a photogrammetric plate. Even using sophisticated image processing techniques, the scanning system still suffers from slow operating speed, intensive programming requirements and not being capable of adapting to different types of target. In order to minimise these deficiencies, an artificial neural network approach has been employed to develop the automatic scanning system. This paper describes the progress made on this application of artificial neural networks.
© (1990) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
W. C. Chiu, W. C. Chiu, E. L. Hines, E. L. Hines, C. Forno, C. Forno, R. Hunt, R. Hunt, S. Oldfield, S. Oldfield, } "Artificial neural networks for photogrammetric target processing", Proc. SPIE 1395, Close-Range Photogrammetry Meets Machine Vision, 13952U (1 August 1990); doi: 10.1117/12.2294345; https://doi.org/10.1117/12.2294345


Back to Top