Cattle identification has recently played an influential role towards understanding disease trajectory, vaccination, production management, and animal ownership assignment. Cattle identification and tracking is defined as the process of accurately recognizing individual cattle and their products via a unique identifier or marker. In the past, some traditional methods such as ear tags, branding, tattooing, and electrical methods have been used frequently; however, their performance is limited due to their vulnerability to losses, duplications, fraud, and security challenges. Moreover, continually using these methods can damage animal’s tissue and therefore cause pain. In long term, it can affect the animal’s appearance, social interaction, other behaviors and ultimately its survival.
To solve the drawbacks of traditional methods, biometric methods have therefore been developed to recognize animals based on physical characteristics or behavioral signs. Some highlighted characteristics of the new method are uniqueness, immutability, and low costs. In additional, biometric methods are non-harmful, do not cause pain and change the appearance of animal. Therefore, animal’s behavior and survivability won’t be affected. Among those methods, muzzle print, retina scan and iris recognition are three biometric modalities that are getting the most attention in the market.
According to Baranov et al (1993), an animal’s muzzle print has capability of identifying an animal with high accuracy as fingerprint of human. Therefore, this can be considered as an accurate and time – immutable biometric identifier. Collecting muzzles print images can be achieved via one of two methods: using ink or a digital camera (live-captured images). The advantage of this method is quite cheap and simple. However, in real-world tests, it took time and the accuracy result was not very impressive also, 94% as reported by MSVM which was published on the International Journal of Image Mining.
Whereas, Retinal vascular patterns (RVPs) in which features are similar to human retinal scans provide considerable cattle identification accuracy over muzzle print. According to “An Evaluation of Retinal Imaging Technology for 4-H Beef and Sheep Identification” study, it is concluded that the false match and false non-match rates of visual verification of retinal images were lower than the rates for nose prints which makes RVPs a more favorable option. Furthermore, retinal patterns can be found in almost all animals; hence this method may be widely applicable in the animal identification market. However, it still remains some disadvantages including acceptability, collectability, and processing time. In additional, the difficulty of capturing a retinal image due to eye diseases and the failure to control the animal’s movement long enough to accurately capture a retinal image are also very challenging issues.
Iris recognition which is lately deployed in many applications is also considered as a great alternative solution for cattle ID system. According to Musgrave & Cambier (2002), due to its effectiveness, iris biometrics has been developed to apply in animal identification. As a study done by group of researchers “A new cow identification system based on iris analysis and recognition”, correct recognition rate of iris identification is 98.33%. Furthermore, iris scanning can be performed briskly and images can be captured digitally. Specifically, different images of animal eyes are taken from different angles from various stationary fixed points cameras. Among these, the best eye image is selected for iris segmentation for further processing which would lead to iris recognition in a meaningful way. With these advantages, iris biometrics has proved to be a very promising solution for animal identification. Currently, it is being used to identify cow and link them to owners in India. “The solution will enable farmers to get unique Ids for their cattle and will help them in insurance claims for livestock”, says the supplier.