|
|||||||||||
Find out more about how CACI MTL can help you!
ARTEMIS Image Mining for Cancers searches electronic image files to find an object or feature of interest (e.g., a needle in a haystack).
IM Applications
- Searching large numbers of image files for faces, fingerprints, DNA patterns, hidden weapons, illegal crops, water pollution, oil/gas deposits, cancerous tumors, etc.
- Searching real time, biometric images for identity verification, fugitive identification, etc.

CACI MTL's ARTEMIS IM is Unique
- Advances the state-of-the-art in search speed, accuracy and versatility
- Uses CACI MTL's proven TWIST/SSP technology to holistically capture a all relevant image features with dramatically less processing time and less storage than traditional techniques
- Automated process eliminates need for manual identification of image features
- Accurate, even with partial, low resolution or low quality images
- Versatile - searches images from multiple types of sensors for a single search goal (data fusion)
- Low cost, PC-based system allows real time searches of local or remote databases
| ARTEMIS Features | ARTEMIS Benefits |
|---|---|
| Inherent Speed - ARTEMIS performs brute force searches much faster than traditional techniques, which extract individual features (eyes, nose, mouth, facial hair, etc.) and perform more complex processing (splines, genetic algorithms, neural networks, etc.). By executing its algorithms on a hardware accelerator chip in a 2.2 Ghz dual processor PC, ARTEMIS can potentially search over 1 million images/sec. | High speed and throughput make real time (on-the-fly) image mining practical for both local and remote database searches |
| Throughput - using intelligent, indexed search techniques and fast PC hardware ARTEMIS can potentially search over 4 million images/sec. | |
| Accuracy - test results to date on small databases of frontal images with nominal lighting and with/without glasses yielded 100% matching accuracy with no false matches. Robust nature of TWIST/SSP technology promises minimal accuracy degradation for low lighting and profile images (up to 30 deg. aspect angle). | High accuracy of image matching enhances effectiveness of operator/analyst |
| Storage Requirements - ARTEMIS' statistical model of an image requires orders of magnitude less storage than the original image | Efficient storage benefits both large, centralized databases and smaller, local databases |
| Input Data - data from multiple types of sensors can support a single search, including biometric, photographic, video, infrared, etc. Sensors can be at different viewing angles and distances from the object. | Versatility allows the fusing (combining) of data from different sensors, viewing angles and distances to enhance search effectiveness |

ARTEMIS Hardware-Software Configuration - PC-based under any Windows O/S, written in Visual Basic
Original Sponsor - CACI MTL developed the original IM technology under IR&D and applied it for the Air Force Research Lab, Rome, NY and for a classified U.S. government customer
Download
brochure
PowerPoint - 629 KB

