Advanced

Convolutional neural network-based cow interaction watchdog

Ardö, Hakan LU ; Guzhva, Oleksiy; Nilsson, Mikael LU and Herlin, Anders H. (2018) In IET Computer Vision 12(2). p.171-177
Abstract

In the field of applied animal behaviour, video recordings of a scene of interest are often made and then evaluated by experts. This evaluation is based on different criteria (number of animals present, an occurrence of certain interactions, the proximity between animals and so forth) and aims to filter out video sequences that contain irrelevant information. However, such task requires a tremendous amount of time and resources, making manual approach ineffective. To reduce the amount of time the experts spend on watching the uninteresting video, this study introduces an automated watchdog system that can discard some of the recorded video material based on user-defined criteria. A pilot study on cows was made where a convolutional... (More)

In the field of applied animal behaviour, video recordings of a scene of interest are often made and then evaluated by experts. This evaluation is based on different criteria (number of animals present, an occurrence of certain interactions, the proximity between animals and so forth) and aims to filter out video sequences that contain irrelevant information. However, such task requires a tremendous amount of time and resources, making manual approach ineffective. To reduce the amount of time the experts spend on watching the uninteresting video, this study introduces an automated watchdog system that can discard some of the recorded video material based on user-defined criteria. A pilot study on cows was made where a convolutional neural network detector was used to detect and count the number of cows in the scene as well as include distances and interactions between cows as filtering criteria. This approach removed 38% (50% for additional filter parameters) of the recordings while only losing 1% (4%) of the potentially interesting video frames.

(Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
IET Computer Vision
volume
12
issue
2
pages
7 pages
external identifiers
  • scopus:85042857889
ISSN
1751-9632
DOI
10.1049/iet-cvi.2017.0077
language
English
LU publication?
yes
id
407c49c4-76f1-4704-a922-18fa70829808
date added to LUP
2018-03-16 09:59:30
date last changed
2018-05-29 11:55:45
@article{407c49c4-76f1-4704-a922-18fa70829808,
  abstract     = {<p>In the field of applied animal behaviour, video recordings of a scene of interest are often made and then evaluated by experts. This evaluation is based on different criteria (number of animals present, an occurrence of certain interactions, the proximity between animals and so forth) and aims to filter out video sequences that contain irrelevant information. However, such task requires a tremendous amount of time and resources, making manual approach ineffective. To reduce the amount of time the experts spend on watching the uninteresting video, this study introduces an automated watchdog system that can discard some of the recorded video material based on user-defined criteria. A pilot study on cows was made where a convolutional neural network detector was used to detect and count the number of cows in the scene as well as include distances and interactions between cows as filtering criteria. This approach removed 38% (50% for additional filter parameters) of the recordings while only losing 1% (4%) of the potentially interesting video frames.</p>},
  author       = {Ardö, Hakan and Guzhva, Oleksiy and Nilsson, Mikael and Herlin, Anders H.},
  issn         = {1751-9632},
  language     = {eng},
  month        = {03},
  number       = {2},
  pages        = {171--177},
  series       = {IET Computer Vision},
  title        = {Convolutional neural network-based cow interaction watchdog},
  url          = {http://dx.doi.org/10.1049/iet-cvi.2017.0077},
  volume       = {12},
  year         = {2018},
}