Sow estrus checking method and estrus checking robot

Abstract

The invention discloses a sow estrus checking method and a sow estrus checking robot, the method applies the sow estrus checking robot, the sow estrus checking robot comprises an image collector and a body temperature measurer, and the method comprises the following steps: collecting a sow image through the image collector, importing the collected sow image into a pre-trained convolutional neural network to carry out convolution operation, identifying an oestrus feature part image of the sow; enabling a body temperature measurer to locate the position of the oestrus characteristic part image, and collecting the temperature information of the oestrus characteristic part of the sow through the body temperature measurer; predicting and identifying the oestrus key state of the sow according to the oestrus feature part image of the sow by utilizing regression analysis to obtain the oestrus key state information of the sow; and performing normalized data processing on the characteristic part temperature information and the oestrus key state information of the sow, and importing the information into a classifier for classification and identification so as to identify the oestrus state of the sow. The method can automatically check the estrus of the sow and judge whether the sow is in an estrus state or not.

Description

technical field [0001] The invention relates to the technical field of a sow estrus checking method and a device, in particular to a sow estrous checking method and a estrous checking robot. Background technique [0002] A sow's estrus cycle usually includes pre-estrus, estrus, late estrus and rest period. During the breeding process, check and grasp the estrus situation of sows in time to ensure that sows are fertilized at a reasonable time, which can effectively improve The pregnancy and farrowing efficiency of sows maximize the benefits of pig raising. [0003] The current sow estrus detection work mainly relies on manual experience judgment and analysis. In the process of large-scale breeding, there are usually many pigs in the pig house. The workload of sows in estrus is relatively large and difficult; [0004] If the sows are artificially induced before the estrus check, the sows can be promoted to be in estrus more effectively, and then the sow's estrus can be judge...

AI Technical Summary

Problems solved by technology

[0003] The current sow estrus detection work mainly relies on manual experience judgment and analysis. In the process of large-scale breeding, there are usually many pigs in the pig house. The workload of sows in estrus is relatively large and difficult;

Images