Research And Implement Of Broilers’ Cough Monitoring System Based On Voice Recognition Technology

China is the world’s second-largest producer of chicken meat and products,with a continuous average growth rate of 5%to 6%per year.The consumption of chicken meat products falls short just after pork.Healthy breeding of broilers is directly related to Food safety and security.With the development of intensive farming,the problems of broiler welfare and management need to be addressed as these could lead to significant losses in broiler production.The common health-related issue in broilers is the respiratory disease which often has a high incidence rate in chicken production with a cough is its primary symptom.Continuous observation is currently the primary methodology of health information acquisition.However,it suffers from high subjectivity,high labor costs and lack of real-time detection.Therefore timely detection of cough status in real-time can be of paramount importance in accessing the health situation in broilers,thus,beneficial to both chicken welfare and economy.Voice recognition technology has been previously applied to monitor the welfare of livestock and poultry,through the sound sample signal to obtain information on the health status of animals and poultry.In this paper,a broiler cough monitoring system based on speech recognition technology.This proposed system was developed based on the requirements of the broiler breeding environment,which provided the basis for the early detection and treatment of broiler respiratory diseases.The system is mainly developed from the following aspects:Firstly,to understand and analyze the development status of speech recognition technology at home and abroad and its application research status in livestock and poultry welfare breeding.Based on the domestic broiler chicken breeding level,the research objectives and specific technical routes of this article were proposed.A broiler sound acquisition device and a wireless sensor network were constructed,using a wireless multimedia sensor technology.To obtain a fully automatic system the proposed system employed the use of network architecture,a sensor node,and a design gateway node.The research on the identification of the snore cough sound and the implementation of the broiler snore cough monitoring system lay the hardware foundation.The broiler sound samples were collected and pre-processed,windowed,and the density of the power spectrum was analyzed.Based on this,an efficient,simple,goal-optimized algorithm that can efficiently be trained on broiler swine coughs sound signal feature extraction algorithm was designed.Based on the fuzzy C-means clustering algorithm,the sound recognition algorithm for broiler cries and coughs was designed with the goal of to achieve an accurate broiler sound recognition.Design and develop a broiler cough monitoring system,which includes an Android-based broiler information integrated management terminal and a WEB-based monitoring platform to enable farmers to obtain and query broiler status information promptly.The experiment was carried out at the Wenye Group’s animal husbandry right-worshipped pepper and broiler chicken breeding center in Bozhou City.20 broilers of 40 days old were selected as the test subjects.The results show that the proposed system was stable,automatic and fast Additionally,the system could identify and classify different broiler coughs the overall cough recognition accuracy was 82.1%.It was concluded that this monitoring system meets the broiler production requirements and provides a monitoring platform for automatic identification of broiler respiratory diseases.