DETEPRE

Determination of reproductive parameters in fish by computer vision

Project description

Many of the European fishery resources are currently overexploited. The resources of interest to the Galician fleet are no exception, and in this sense some recovery plans have been proposed by the EU. In the Precautionary Approach to fisheries management is intended to conserve the sufficient Reproductive Potential of a Stock (PRS) that allows the sustainable exploitation of the same. For this reason, the need to calculate the PRS, which requires knowledge of population size, age composition, proportion of mature individuals, their condition and their fecundity, is increasingly evident. However, despite their recognized importance for the correct population assessment, estimates of the proportion of mature individuals and fertility are rarely performed, due to the complexity of the analysis, involving the use of techniques and protocols with a high Economic and human cost. This project aims to develop methods and tools that facilitate and automate the evaluation of fish reproductive parameters and, ultimately, the reproductive potential of their populations. The project intends to use the new technologies to solve one of the bottlenecks that has the analysis of the dynamics of fish stocks: to have accurate estimates of the reproductive potential of the populations. This calculation requires, among other variables, the proportion of mature individuals in the population and the fecundity at each age, the two parameters that are the subject of this project.

This project was structured around three main objectives:

Study of the regression processes of post-ovulatory follicles (FPO) and beginning of maturation.Development and application of new histological and microscopy tools that allow the rapid diagnosis of the maturation stage and the putting activity of the individual.

Design of an automatic system, based on computer vision, that operates on digital images of histological preparations of ovaries.
It was found that precision can be increased in the calculation of fertility using automated computing tools based on a semi-automatic processing of the histological images of fish gonads, while decreasing the working time of the researcher or technician, Due to the reduction of manual labor and the automation of the routine process. Has developed software that will allow the Govocitos not only estimating fertility by applying the method esterológico quickly and accurately in many species, but it also possible to study the dynamics of recruitment and oocyte development. Added values ​​for the use of the tools of the new technologies can be listed: accessibility to data (including images) and results through a centralized database with adaptable access interfaces; Reproducibility of data interpretation and manual tasks; Statistical analysis of the data with possible direct storage of the results in the database; Remote work without and with live connection to the system; Increased precision of the experiments. All this using a database that allows information sharing and online work for researchers. The software has been presented in various forums and has received great acceptance and demand from the scientific community, so it will be distributed under free software license in early 2013.

Associated events

More projects