SAREO - Synthetic Aperture Radar and Electro-Optical Imagery
Contracted awarded by the Defence Research and Development Canada-Ottawa to investigate the evaluation methods and analysis of segmentation algorithms for Synthetic Aperture Radar (SAR) and Electro-Optical (EO) Imagery (SAREO)
Image segmentation is a fundamental process for the automation of SAR and EO image analysis. The application of a segmentation algorithm to a scene divides an image into regions for which a specific parameter is internally consistent. Such scene segmentation can be used as a basis for feature extraction, change detection and target detection. Segmentation is also a crucial part of the Automatic Target Recognition (ATR) process for imagery, whereby information pertaining to a single target is extracted and separated from the rest of the image.
For this contract, Array Systems performed the following research:
- Investigated the literature for evaluation methods of image segmentation and recombination applications.
- Designed and standardized performance measures baselines along with appropriate data sets for segmentation span
- Established and documented an appropriate evaluation procedure.
- Generated and compared evaluations of several current segmentation methods.
The Radar Data Exploitation (RDE) group at DRDC-Ottawa is responsible for maintaining and improving the Canadian Forces scientific and technological capability to analyze, interpret and exploit signal and imagery data from radar remote sensors. To enable this, RDE is conducting research into several leading edge technologies to support the development of applications for target detection and recognition from radar imagery and multi-sensor data fusion of images. RDE uses Array's analysis on image segmentation evaluation methods as part of their research.