Browsing by Author "Le Roux, Solomon Petrus"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Animal-borne behaviour classification for sheep (Dohne Merino) and rhinoceros (Ceratotherium simum and diceros bicornis)
    (BioMed Central, 2017-11-21) Le Roux, Solomon Petrus; Marias, Jacques; Wolhuter, Riaan; Niesler, Thomas
    Background: The ability to study animal behaviour is important in many fields of science, including biology, behavioural ecology and conservation. Behavioural information is usually obtained by attaching an electronic tag to the animal and later retrieving it to download the measured data. We present an animal-borne behaviour classification system, which captures and automatically classifies three-dimensional accelerometer data in real time. All computations occur on specially designed biotelemetry tags while attached to the animal. This allows the probable behaviour to be transmitted continuously, thereby providing an enhanced level of detail and immediacy. Results: The performance of the animal-borne automatic behaviour classification system is presented for sheep and rhinoceros. For sheep, a classification accuracy of 82.40% is achieved among five behavioural classes (standing, walking, grazing, running and lying down). For rhinoceros, an accuracy of 96.10% is achieved among three behavioural classes (standing, walking and lying down). The estimated behaviour was established approximately every 5.3 s for sheep and 6.5 s for rhinoceros. Conclusions: We demonstrate that accurate on-animal real-time behaviour classification is possible by successful design, implementation and deployed on sheep and rhinoceros. Since the bandwidth required to transmit the behaviour class is lower than that which would be required to transmit the accelerometer measurements themselves, this system is better suited to low-power and error-prone data communication channels that may be expected in the animals habitat.
  • Thumbnail Image
    Item
    A prototype animal borne behaviour monitoring system
    (Stellenbosch : Stellenbosch University, 2016-03) Le Roux, Solomon Petrus; Wolhuter, R.; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: Rhinos across the globe are suffering from an immense onslaught from rhino poachers. The latter show no regard for the remaining rhino species. New technologies must be implemented to provide scientists, biologists and nature conservationists with key information regarding the behaviour and well being of animals. This project entailed the design and development of a prototype Animal Borne Behaviour Monitoring System (ABBMS). The system was based on animal borne sensor driven devices known as WildMotes. The latter were used to collect and communicate sensor data to a base station, by means of a multi-hop Wireless Sensor Network (WSN) that exists between WildMotes, repeaters and the base station. The project considered the hardware design of the WildMotes, which included various components such as an ultra-low power microcontroller, GPS, accelerometer, nano-power tilt and vibration sensor, RF communication module, microSD card and FRAM modules. Apart from the hardware design, the project included all software required of the ABBMS. Initially the WildMotes collected data that could be used for the automatic behaviour classification of animals, by means of computer based techniques. The data was utilised with techniques such as linear- and quadratic discriminant analysis and decision trees, to classify the behaviour of rhinos and sheep with high accuracies. This behaviour included running, walking, standing, grazing and laying down. Furthermore, this project successfully implemented an On-animal Behaviour Classification System (OABCS). To the best of our knowledge, this is the first implementation where behaviour classification is performed in real time, on the animal. This technique provides live updates of animal behaviour, as opposed to post processing, computer based techniques. The OABCS was able to accurately distinguish between similar behavioural classes as above mentioned. In addition to the OABCS, a nano-power tilt and vibration sensor was applied, as an ultra-low power alternative, to classify the behaviour of animals. The latter could accurately distinguish between the same behaviour, while consuming very little energy. This technique was further utilised, in combination with the OABCS, to extend the battery life of the WildMotes from roughly 47 days to 270 days. Finally, GPS coordinates were obtained and utilised to reveal repetitive movement patterns of rhinos, by means of a heat map. In future work, the ABBMS can be combined with the OABCS, GPS locations and key stress level indicators, such as pulse rates, to learn more about endangered species and serve as tools in the fight against illegal poaching activities.