Automatic System for Zebrafish Counting in Fish Facility Tanks
In this project we propose a computer vision method, based on background subtraction, to estimate the number of zebrafish inside a tank. We addressed questions related to the best choice of parameters to run the algorithm, namely the threshold blob area f
- PDF / 1,465,229 Bytes
- 9 Pages / 439.37 x 666.142 pts Page_size
- 8 Downloads / 185 Views
ESC-ID Lisboa, Instituto Superior T´ecnico, Universidade de Lisboa, Lisbon, Portugal [email protected], [email protected] 2 Champalimaud Centre for the Unknown, Lisbon, Portugal {ana.certal,carlos.maodeferro,joana.monteiro,jose.cruz, ricardo.ribeiro}@neuro.fchampalimaud.org
Abstract. In this project we propose a computer vision method, based on background subtraction, to estimate the number of zebrafish inside a tank. We addressed questions related to the best choice of parameters to run the algorithm, namely the threshold blob area for fish detection and the reference area from which a blob area in a threshed frame may be considered as one or multiple fish. Empirical results obtained after several tests show that the method can successfully estimate, within a margin of error, the number of zebrafish (fries or adults) inside fish tanks proving that adaptive background subtraction is extremely effective for blob isolation and fish counting. Keywords: Computer vision · Zebrafish counting subtraction · Hu moments · Image processing
1
·
Background
Introduction
Zebrafish (danio rerio) is a small freshwater fish that is widely used as an animal model in biomedical research [2]. Research laboratories around the world require a huge number of individuals to perform a great variety of experiments. Those fish are breed and maintained in big fish facilities managing hundreds to thousands fish tanks. Usually these tanks are standardized containers (from 3 to 8 L for instance) which may host up to dozens of animals each. Obtaining an up to date count of the total number of animals in a fish facility is an essential task, performed by human technicians who manually extract animals with the help of small fish nets. This manual counting process requires a significant amount of time and is error prone. Moreover, handling animals for counting induces significant stress, with all the harmful consequences that may cause to the animals and, consequently, affect the scientific experiments they are involved in. c Springer International Publishing Switzerland 2016 A. Campilho and F. Karray (Eds.): ICIAR 2016, LNCS 9730, pp. 774–782, 2016. DOI: 10.1007/978-3-319-41501-7 86
Automatic System for Zebrafish Counting in Fish Facility Tanks
775
Finding a noninvasive automatic procedure to obtain the precise number of zebrafish in facilities tanks, avoiding all the disadvantages of manual counting, is a long sought goal of fish facilities managers. Today there are many examples of complex applications of computer vision techniques such as: optical character recognition, machine inspection, 3D model building, medical imaging, face detection, visual authentication and people tracking [12]. Some of the previous applications make use of techniques such as background subtraction which have particular interest in this article. Background subtraction is specially relevant when, for example, the need for isolating moving regions in a sequence of images arises. In fish counting, since images are two-dimensional, the main difficulties to overcome are: regions
Data Loading...
