Animal running monitoring system and running state recognition method based on neural network

Abstract

The invention discloses an animal running monitoring system and a running state recognition method based on a neural network. The animal running monitoring system comprises an acquisition device and an intelligent terminal application, and the acquisition device acquires contact pressure, heart rate and acceleration information of monitored animals and judges whether effective contact occurs or not according to the magnitude of the contact pressure; the intelligent terminal application receives effective animal data of the acquisition device and recognizes the effective animal data into the animal running state. By means of the recognition method, based on a neural network model, a sample set can be trained and learned, and based on the trained model, the running state of the animals can be recognized. By means of the system, animal running can be automatically detected and remotely monitored in real time, manual detection or recording is not needed, and a lot of manpower can be saved;by means of the method, the difference like the type, the age, the individual and the motion state of the animals can be eliminated, all monitored objects are truly customized, and the purpose of precisely and effectively detecting whether the animals are running or not is achieved.

Description

technical field [0001] The invention relates to the field of animal health monitoring, in particular to an animal running monitoring system and a neural network-based running state recognition method. Background technique [0002] Running status is an important health indicator of animals. Running monitoring is one of the important ways to evaluate the health status of animals. The timeliness, accuracy and reliability of running monitoring directly affect the prevention, diagnosis, treatment and postoperative recovery of diseases. [0003] Traditional animal running detection uses a single running sample point, which is very inconvenient to track changes in animal running. At present, the measurement of animal running is mainly done manually. The owner needs to regularly check the daily status of animals every day, manually measure animal running, record running, and draw Running curves, that's a lot of work. Traditional running detection methods cannot implement high-frequ...

AI Technical Summary

Problems solved by technology

[0003] Traditional animal running detection uses a single running sample point, which is very inconvenient to track changes in animal running. At present, the measurement of animal running is mainly done manually. The owner needs to regularly check the daily status of animals every day, manually measure animal running, record running, and draw Run the curve, it's a lot of work

Traditional running detection methods cannot implement high-frequency detection

[0004] In addition, different kinds of animals, different age groups of the same kind of animals, different individuals of the same kind of animals of the same age, and different running states of the same individual animals often have different heart rates. For example, the heart rate of an adult dog is 68-80 beats in a non-running state / min, the normal heart rate of puppies is 80-120 beats / min, therefore, animal heart rate monitoring should take into account differences in animal species, age, individual, exercise status, etc., and traditional methods are difficult to fully achieve

Images