Application Of Dairy Cow Body Condition Scoring Based On Machine Vision

Dairy cow body condition scoring (BCS) is one of the most common processing in modern aquaculture production. In order to obtain a stable quality of agricultural products and improve the efficiency of production management, we hope the process of breeding can be automated and product yield and quality can be optimized. In this paper, the design system for assessing dairy cow body condition which based on machine vision can increase dairy production and improve production efficiency, and produce more high quality dairy and improve workers’environment is of great significance.Dairy cow body condition is the study object of this paper,on the basis of existing theories’ summary and the result of this area study, we propose a design which use Kinect Sensor device, and complete dairy cow body condition scoring system design; develop a hardware and software platform, the hardware and software platform design to achieve simulation assessment experiment based on machine vision technology, compliance with the manual assessment conclusions. The paper though the following aspects of works study this research inscription:Using Kinect Sensor device shoot on the object and generate speckle images based on lighting encoding technology. Infrared transmitter and infrared camera which can combine to capture image form a three-dimensional shape of the object scene. Using this technology to extract the feature points for image segmentation. Then towards the segmented images we extract and match its features.Finally, we are able to realize the cow surface contour description. The processing can complete the assessment of dairy cow body condition.On the basis of system requirements analysis, the paper develop a design which is based on Kincet Sensor Intel(?) AtomTM processor-based embedded platform and the Windows operating system, using the Kinect Sensor image acquisition sensors and other hardware devices. Kinect Sensor collects the depth images as the source image target. Using OpenCV2.3.1library as the basis of pattern recognition processing based on machine vision and image processing technology; build an automated dairy yield evaluation system. The target images can be gathering from three different angles, analysis and assessment achieve cow body condition scoring. This design can replace the existing, very time-consuming and error greater manual scoring method. The system analysis results can help to develop dairy cow breeding program. The value of this system can help save a lot of manpower and dairy producing time.The study towards dairy cow body condition assessment have very broad application prospect.1. Dairy cow body condition measurement study history and present situation.2. Using Key technologies in Dairy cow body condition assessment system.3. Using the depth images assessment methods and principles.4. Using the depth images assessment implementation.5. Experiment results and analysis.6. Summary and prospect.