Predicting growing -finishing swine’s weight with image analysis

• Grant number: 14/18458-6
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: December 01, 2014
• End date: February 29, 2016
• Field of knowledge: Agronomical Sciences - Agricultural Engineering - Rural Buildings and Ambience
• Principal Investigator: Késia Oliveira da Silva Miranda
• Grantee: Isabella Cardoso Ferreira da Silva Condotta
• Host Institution: Escola Superior de Agricultura Luiz de Queiroz (ESALQ). Universidade de São Paulo (USP). Piracicaba , SP, Brazil
• Associated scholarship(s): 15/07254-3 - Images analysis in prediction of swine weight, BE.EP.IC

Abstract

Food production is increasingly focused on the quality of the final product and its effects on the environment, as well perceives arising demand for products from animals raised respecting the precepts of welfare quality. Bearing in mind that the intensive system of creation has many sources of stress, it is necessary to search for alternatives that promote an improvement animal well-being without an major increase in labor. The constant human attendance on the animal management activities, besides increasing the cost of production, becomes a source of animal stress. The development of methods for monitoring the physical conditions of animals from distance and in real time, in addition to the distance control of the installations allow minimal human intervention on their behavior. This process can decrease sources of stress through the production cycle, improve quality levels and feed conversion, improve the quality of data obtained, open new markets and fortifying the production and exports of pork in Brazil. The current payment equation of typified carcasses requires that the live pig weight be known, thus there is a need of weighing pigs for commercialization. Several researchers have successfully shown that swine dimensions extracted from digital images can be used to predict weight within approximately 5% margin of error. However, there are difficulties with this approach that makes it very difficult to automate. Alternatively, the use of the 3D camera eliminates the problems of the first approach. Therefore, this project aims, in partnership with the Meat Animal Research Center (MARC) from Nebraska, to extract pigs weights data from correlation with 3D images.