High-Speed Tube Diameter Measurement Using GPU Thread-Level Parallelism
This paper presents a GPU-based acceleration method for tube diameter measurement in gastrointestinal medical images. An exhaustive image measurement algorithm is mapped onto the GPU thread architecture to enable massive parallel processing. Experimental results show that the proposed method reduces analysis time by approximately 99% compared with a CPU-based implementation.
Authors
- Tomoyuki Yokogawa (ORCID: https://orcid.org/0000-0001-6681-2608)
- Yoichiro Sato
- Mitsuru UEDA
- Masafumi Kondo
Institutions
- Okayama University of Science (JP)
- Okayama Prefectural University (JP)
Publication Details
- Journal
- IEEJ Transactions on Electronics Information and Systems
- Published
- 2026-06-30
- DOI
- https://doi.org/10.1541/ieejeiss.146.679
- Primary Topic
- Parallel Computing and Optimization Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00