A Locally Adaptive Method for Boundary Detection
- Mission Support and Test Services, LLC, Nevada National Security Site
Edge detection techniques predict boundaries between objects or regions in an image. Many classical methods rely on sharp changes in luminance, chromaticity, or texture, and images without these features are difficult to partition correctly. We present a supervised boundary detection method named Locally Adaptive Discriminant Analysis (LADA) that finds edges in images containing high noise and gradual changes between classes. The user trains the algorithm by providing locations of known materials within the image, but no assumptions are made on the material geometries in the image. The algorithm then uses statistics of anisotropic local data to place each pixel in the most probable class. The algorithm’s success is demonstrated on an optical image of a laser-induced shockwave.
- Research Organization:
- Nevada National Security Site (NNSS), North Las Vegas, NV (United States)
- Sponsoring Organization:
- USDOE National Nuclear Security Administration (NNSA)
- DOE Contract Number:
- DE-NA0003624
- OSTI ID:
- 1471704
- Report Number(s):
- DOE/NV/03624-0071
- Resource Relation:
- Conference: Conference on Data Analytics , Santa Fe, New Mexico, March 7, 2018
- Country of Publication:
- United States
- Language:
- English
Similar Records
A Locally Adapting Technique for Edge Detection using Image Segmentation
Computer-aided detection of interstitial abnormalities in chest radiographs using a reference standard based on computed tomography