ITEE Innovation Expo 2008 : Project Details

Detecting Fertility in Cattle Using Image Processing

Abstract Project title: Detecting fertility in cattle using image processing Fertility of dairy farm cattle can be evaluated using photographs of a heat detection device normally placed in cows¡¯ back. The device that with a blue or red dye packs in the middle and it turns red or blue (depending on the ink colour used) when the cow is in its fertile period and subsequently ready for artificial insemination. This project aims to build a fast and simple algorithm to analyse the colour change in the detection device ultimately eliminating the need for imagery. This analysis is based on digital image recognition and processing of which there are two methods: 1) Colour-based segmentation, which identifies different colours in image by analyzing the L*a*b* colorspace; 2) Template matching (cross-correlation), which efficiently calculates similarities between the template and partial images of various width and heights; 3) Colour-based segmentation, which uses to compare the characteristic of rgb value of different pixels, thus to distinguish the colour These methods, however have their advantages and disadvantages and the third segmentation method get better result.