An automatic colour-based computer vision algorithm for tracking the position of piglets

Abstract

Artificial vision is a powerful observation tool for research in the field of livestock production. So, based on the search and recognition of colour spots in images, a digital image processing system which permits the detection of the position of piglets in a farrowing pen, was developed. To this end, 24,000 images were captured over five takes (days), with a five-second interval between every other image. The nine piglets in a litter were marked on their backs and sides with different coloured spray paints each one, placed at a considerable distance on the RGB space. The programme requires the user to introduce the colour patterns to be found, and the output is an ASCII file with the positions (column X, line Y) for each of these marks within the image analysed. This information may be extremely useful for further applications in the study of animal behaviour and welfare parameters (huddling, activity, suckling, etc.). The software programme initially segments the image in the RGB colour space to separate the colour marks from the rest of the image, and then recognises the colour patterns, using another colour space [B/(R+G+B), (G-R), (B-G)] more suitable for this purpose. This additional colour space was obtained testing different colour combinations derived from R, G and B. The statistical evaluation of the programme's performance revealed an overall 72.5% in piglet detection, 89.1% of this total being correctly detected.Additional key words: animal behaviour, image analysis, pattern recognition, piglets activity, segmentation. ResumenSistema de visión artificial basado en el color, para el seguimiento de la posición de lechones La visión artificial es una potente herramienta de observación al servicio de la investigación en el campo de la producción ganadera. En este sentido, se ha desarrollado un software de tratamiento digital de imagen, basado en la búsqueda y reconocimiento de manchas de color en la imagen, que permite la detección de la posición de los lechones en una corralina de maternidad. Para ello, se capturaron 24.000 imágenes en cinco tomas (días), con un intervalo de 5 segundos entre cada dos imágenes. Los nueve lechones presentes en estas camadas fueron pintados en el lomo y laterales con colores distintos y alejados entre sí en el espacio RGB. El programa precisa la introducción por el usuario de los patrones de color a buscar, devolviendo como salida un fichero ASCII con las posiciones (columna X, fila Y) de cada una de esas manchas en cada imagen analizada. Esta información podría ser de gran utilidad en posteriores aplicaciones para el estudio de paráme-tros de comportamiento y bienestar de los animales (hacinamiento, actividad, amamantamiento, etc.).El software realiza una segmentación previa de la imagen en el espacio RGB, para aislar las manchas de color del resto de la imagen, y posteriormente el reconocimiento de los patrones de color utilizando otro espacio de color [B/(R+G+B), (G-R), (B-G)] más adecuado para tal fin, obtenido a partir de pruebas con diferentes combinaciones ...

Foreground detection of group-housed pigs based on the combination of Mixture of Gaussians using prediction mechanism and threshold segmentation

“…Among these, research on the accuracy of foreground detection of pigs is basic work for advancing subsequent research (Kashiha et al, 2013). Foreground detection of individual pigs is mentioned to varying degrees in some literature (Ahrendt, Gregersen, & Karstoft, 2011;Hu & Xin, 2000;Lind, Vinther, Hemmingsen, Hansen, 2005;McFarlane & Schofield, 1995;Navarro-Jover, et al, 2009;Wang, Yang, Winter, Walker, 2008). In that literature, however, foreground detection technology is not the focus; rather, it is used to address requirements of specific applications.…”

“…Some current computer vision systems that are used for video surveillance of group-housed pigs require that the pigs are marked (Navarro-Jover et al, 2009). Tu, Karstoft, Pedersen, and Jørgensen (2013) proposed a foreground detection algorithm based on loopy belief propagation.…”

Segmentation of sows in farrowing pens

“…In the past 20 years, a few algorithms for detecting and tracking pigs are introduced in [15][16][17][18][19][20][21][22][23][24]. But most of the algorithms have not discussed their results in the complicated scenes, such as light changes and motionless foreground objects as well as dynamic background.…”

Video multitracking of fish behaviour: a synthesis and future perspectives

“…2009; Blaser et al. 2010), animal well‐being studies (Navarro‐Jover et al. 2009; Winandy and Denoël 2011), behavioural ontogeny (Fontaine et al.…”

“…It is an automatized procedure that determines animal position over time and gives the resulting tracks with a large array of data such as distance travelled, speed or space used (Noldus et al 2001;Maggio and Cavallaro 2011). The quantitative approach of this methodology has made it possible to collect straightforward data in fields as varied as ecotoxicology (Kanen et al 2005;Jakka et al 2007;Denoël et al 2010), neurotoxicology (Eddins et al 2010), behavioural brain research (Mathur et al 2010;Silverman et al 2010), pharmacology (Pinhasov et al 2005;Singh et al 2009;Cachat et al 2010), genetic and behavioural screen-ing (Orger et al 2004;Egan et al 2009;Blaser et al 2010), animal well-being studies (Navarro-Jover et al 2009;Winandy and Denoël 2011), behavioural ontogeny (Fontaine et al 2008;Fukuda et al 2010), social behaviour (Newlands and Porcelli 2008;Salierno et al 2008), cognition (Bisazza et al 2010) and ethology and behavioural ecology (Pritchard et al 2001;Speedie and Gerlai 2008).…”