North American Journal of Fisheries Management · 2014 · 16 citations · 10 references
Life SciencesComputer ScienceZoologyComputer VisionFish BiologyMarine ZoologyAnimal TrackingMarine Monitoring SystemsAnimal Movement MonitoringMotion PredictionUnderwater Imaging Systems
Abstract To reduce staffing demands and enhance observational accuracy, we designed, implemented, and tested a computerized system for processing underwater video clips captured by static cameras and removing “quiet” frames in which no activity is detected. When the system detects activity, it tracks and counts the moving object. The implementation uses an adaptive background-subtraction algorithm for detection and motion prediction for tracking. The system can be used to reduce the amount of video that must be reviewed by personnel, and it can produce total fish passage counts through the monitored area. The automated system reduced the total number of video hours requiring review by an average of 87.5% for count window videos and 83.5% for videos of picketed leads and lamprey passage orifices. The software detected 98.6% of the 144 Pacific Lampreys Entosphenus tridentatus that were observed in 185 h of raw video. Because the system is fully automated, monitoring requirements are negligible, and the cost reduction for fish monitoring is proportional to the number of quiet frames removed. Received March 27, 2013; accepted September 23, 2013 Published online January 30, 2014
10
Towards robust automatic traffic scene analysis in real-time
Daniela Koller, Joseph Weber, T. Huang et al. · 2002 · 380 citations
Wenjun Li, Peter W. Sorensen, Daniel D. Gallaher · The Journal of General Physiology · 1995 · 212 citations · Full text
Mary L. Moser, Alicia L. Matter, Lowell C. Stuehrenberg et al. · Hydrobiologia · 2002 · 78 citations
Damon H. Goodman, Stewart B. Reid, Margaret F. Docker et al. · Journal of Fish Biology · 2008 · 61 citations
Life Sciences, Genetics, Zoology +16