Repository hosted by TU Delft Library

Home · Contact · About · Disclaimer ·

Automatic Characterization of Electro-Optical Sensors with Image Processing, using the Triangle Orientation Discrimination (TOD) Method


Author: Lange, D.J. de · Valeton, J.M. · Bijl, P.
Publisher: SPIE
Place: Bellingham, WA.
Institution: TNO Fysisch en Elektronisch Laboratorium TNO Technische Menskunde
Source:Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XI, 26-27 April 2000, Orlando, USA, 104-111
Proceedings of SPIE
Identifier: 95292
doi: doi:10.1117/12.391770
Keywords: Vision · Algorithms · Cameras · Error correction · Image enhancement · Imaging systems · Infrared imaging · Optical sensors · Minimum resolvable contrast (MRC) method · Minimum resolvable temperature difference (MRTD) method · Triangle orientation discrimination (TOD) method · Image sensors · Thermal imaging · Minimum resolvable contrast


The objective characterization of electro-optical sensors and that of image enhancement techniques has always been a difficult task. Up to now the sensor is characterized using the minimum resolvable temperature difference (MRTD) or the minimum resolvable contrast (MRC). The performance of image enhancement techniques is done by visual inspection by a human observer. Since in more and more cameras some kind of image processing is applied, a more elaborate test is needed that can measure the performance of the combination of the sensor and the image processing. A good candidate is the TOD (Triangle Orientation Discrimination) method, which is developed as an alternative for MRTD and the MRC methods. We are investigating how the standard TOD-method can be extended to cameras with image processing and whether these measurements can be automated. An algorithm is under development, which is based on the TOD-method and predicts the characterization by human observers of camera-system performances. The algorithm combines the TOD-method, an early-vision model, and an orientation discriminator. The algorithm uses the same images as used in human-observer experiments. After correction for the physical properties of the display and the human eye, the algorithm tries to find the orientation of the stimulus. The algorithm can also predict the performance of only image processing using a simple scene-generator in stead of a camera setup.