Study On Application Of Image Recognition Technique In The Cow Body Condition Score

The cow body condition score is an important measure of the energy metabolism status and the degree of fat deposition. It is an indispensable tool for the dairy management. Different cows have the different cow body condition scores. The traditional cow body condition score method requires the professionals who master the standard of body condition score. And there are some subjective differences between the scores from different raters.This paper presents an automatic method of cow body condition score using digital image recognition technology by analyzing the relevant standards. And detect the feasibility of the new automatic method on cow body condition score through interrelated experiments.The image recognition is a technology to identify a variety of different models of objectives and targets, by analyzing and understanding the image using computer. Currently in the field of face recognition and facial expression recognition it has become increasingly mature. But there are no related researches on the application of image recognition in the cow body condition score. This paper introduces the digital image recognition to the field of the cow body condition score. The order is to achieve the automatic scoring of the cow body condition.Propose a new method of body condition score using digital image recognition for Chinese Holstein cows. This paper presents an automatic method of cow body condition score using digital image recognition technology by analyzing the relevant standards. And detect the feasibility of the new automatic method on cow body condition score through interrelated experiments.There are three main issues in this study as follow.(1) Location of the cow in the image. To get rid of the background image that is useless for the image recognition.(2) Preprocessing of the black and white. Because the major dairy-based cow is Chinese Holstein cattle (also known as Chinese black and white cows). And it is the current largest number of the most widely distribution, the highest milk yield of dairy cattle breed type. Because the body color is black and white, the black and white is a very big impact for background removal and image recognition。(3) To find a suitable method for cow image recognition through lots of experiments. It is critical to the accuracy of the final result and it will affect the application in practice.(4) Shoot lots of cow’images in different BCS scores, and establish a moderate size cows’image library. When taking the cow images, there are professionals who have rich experience in body condition score scoring the cow. And record them carefully.The research objective is to find a automatic method of cow body condition score using digital image recognition technology. Therefore, first, we must preprocess the black and white. Then, select the appropriate body condition score method for image recognition. Finally, compare the score from automatic score and artificial score.The new method can overcome the shortcoming of the traditional methods. Play a catalytic role on dairy cattle breeding and management.