Research on Dynamic Object Tracking Using River Navigation Images - towards River Traffic -
-
- KONDO Masaki
- 東京海洋大学大学院海洋工学系
-
- INAISHI Masaaki
- 東京海洋大学大学院海洋工学系
-
- FURUYA Tadasuke
- 東京海洋大学大学院海洋工学系
-
- NAKAGAWA Masaki
- 東京農工大学工学部
Bibliographic Information
- Other Title
-
- 河川航行映像を用いた動的物体の追尾と進路予測に関する研究-河川交通に向けて-
Abstract
<p>In this research, we aim to detect ships sailing around our ship from river navigation images using Faster R-CNN and predict their course. We changed the parameters of Faster R-CNN and compared detection rates of ships. We detected navigating ships as obstacles using the best discriminator and river navigation images for evaluation. We predicted the course of the detected navigating ships. The results are summarized as follows:</p><p>(1)We created data set of eleven categories that can be used for machine learning from learning river environment images.</p><p>(2)We detected sailing ships using Faster R-CNN. We examined the detection rate by changing resizer image, anchor size and number of times of learning which are parameters of Faster R-CNN.</p><p>(3)We carried out course prediction from the detected navigating ship's position information. We evaluated the accuracy of predicted course.</p>
Journal
-
- The Journal of Japan Institute of Navigation
-
The Journal of Japan Institute of Navigation 138 (0), 101-109, 2018
Japan Institute of Navigation
- Tweet
Details 詳細情報について
-
- CRID
- 1390845712971788544
-
- NII Article ID
- 130007387326
-
- ISSN
- 21873275
- 03887405
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- Crossref
- CiNii Articles
- KAKEN
-
- Abstract License Flag
- Disallowed