Premium
Evaluation of a Method to Count and Measure Live Salmonids in the Field with a Video Camera and Computer
Author(s) -
Irvine J. R.,
Ward B. R.,
Teti P. A.,
Cousens N. B. F.
Publication year - 1991
Publication title -
north american journal of fisheries management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.587
H-Index - 72
eISSN - 1548-8675
pISSN - 0275-5947
DOI - 10.1577/1548-8675(1991)011<0020:eoamtc>2.3.co;2
Subject(s) - microcomputer , fish <actinopterygii> , computer science , video camera , range (aeronautics) , oncorhynchus , field (mathematics) , computer program , measure (data warehouse) , software , fishery , statistics , computer vision , mathematics , biology , telecommunications , database , engineering , chip , pure mathematics , programming language , aerospace engineering , operating system
A prototype of a computerized video‐camera system successfully counted and measured juvenile salmon in the field. Known numbers of smolts of coho salmon Oncorhynchus kisutch swimming through transparent acrylic tunnels were videotaped, and the videotapes were subsequently interpreted with a microcomputer. Unaided, the computer underestimated the number of fish because fish sometimes overlapped while swimming through the tunnels and were thus recorded as a single fish. However, revised estimates were accurate when portions of the tapes were viewed by an operator and expansion factors for computer‐generated counts determined. When a mouse pointing device was used to obtain measurements from videotapes of individual fish swimming through the tunnels, these measurements were accurate. The software did not reliably produce accurate fish‐length measurements when unassisted, again because of fish crowding and overlap in the tunnels. When a computer operator viewed the tapes and rejected those fish measurements that were obviously in error, there were no differences among computer‐generated length‐frequency distributions and those obtained by direct measurements of the same fish in the field. The technology exists to use computerized video‐camera systems to accurately count and measure live fish when the size range of fish being sampled is small. Potential benefits to the user of this developing technology include time savings, reduction of stress to the fish, the possibility of more accurate and consistent results, and the assurance of permanent records in the form of videotapes and computer files.