Research On The Key Techniques Of The Digital Management Of Dairy Cattle

Agricultural informationalization degree is an important symbol of agricultural modernization, with information technology represented by computer technology as a challenge to traditional agriculture. Along with the rapid rise in China's dairy industry, there have been food safety and breeding problems associated with dairy farming machinery and equipment, dynamic monitoring of physiological indexes of dairy cows, milk quality testing, etc., which is a higher requirements to the status quo of traditional cow breeding. Therefore, digital management technology is an inevitable trend to modern dairy industry development.The identification of individual dairy cattle is the key and basis of digital dairy cattle feeding. In order to improve sensitivity, accuracy and anti-interference in dairy cattle identification, the whole scheme of the dairy cattle identification system was designed on the basis of in-depth researches of the radio-frequency identification technology to realize real-time monitoring and individual cow identification, describes the physical foundations of UHF radio frequency identification system, analyzes UHF radio frequency identification tag performance. Focusing on the existing multi-tag collision of long distance RFID system, two anti-collision algorithms, namely, the binary search algorithm and slot-Aloha algorithm, were compared systematically, and the improved binary search algorithm was selected and finally the perfect effect was obtained. In the data transmission process, the method that the data with the same bits were not transmitted was used in order to reduce the data exchange time, to improve identification efficiency. In this way, the problem of collision about dairy cattle enter into radio-frequency area was settled.Second, from a practical point of view, an automated image acquisition system is established. Cow body linear size can be assessed on the premise that the three-dimensional images be collected clearly and simultaneously so to carry out image processing and identification. The system consists of camera device and positioning equipment, adopts modular design. Computer as a host computer be connected to SCM through RS232 serial port via a dedicated level converter MAX232 chip, controlling and coordinating the functioning and the communication between the modules so that the modules work according to the scheduled timing. SCM as the next bit machine is used to collect data and control gates or railings. Third, in image processing part, color space is studied deeply. That enables image format conversion, image denoising, image enhancement and other pre-processing operation, so as to meet research needs. In image preprocessing algorithms, after comparing a wide range of algorithms, effective methods are adopted according to the actual situation of dairy cows images, thus improves the system's overall performance.Finally, this paper studies the problem of image feature extraction. Feature extraction is the difficulties in reach. Image feature extraction method has a direct impact on image recognition, feature extraction, and cows linear assessing. The analysis of the edge feature extraction method gives way to deal with black and white dairy cattle. On the basis of analyzing the conventional Canny algorithm, this proposed an edge-detection method based on the canny operator.This key technology research of cows serves as the initial completion of the individual cow identification system design, programming and run debugging; in order to achieve individual cow linear assessing, image processing and feature extraction have been studied, and the idea of extracting local feature points is put forward.