Sichuan farmers' "Sanxiangxiang": poultry and livestock recognition help agricultural reform and development

In order to realize the intelligence of farm and poultry, poultry, and livestock breeding industry, help rural reforms, and add youthful vitality to rural revitalization, students from the School of Information Engineering, Sichuan Agricultural University have recently come to the school's livestock and water poultry breeding bases to conduct field investigations, focusing on focusing on water poultry and animal husbandry. Several important indicators such as the level of automation technology and the deployment of artificial intelligence equipment, based on the concept of "digital breeding" breeding, the accurate identification, accurate evaluation and accurate feedback of scientific management data of poultry and animals are implemented.

According to the survey, my country's animal husbandry output value has accounted for 34%of the total agricultural output value, and the output of water poultry breeding accounts for more than 90%of the world's total. With the continuous advancement of the process of socialist modernization in my country, the gap between the industrial production level represented by the existing traditional agricultural operation model is increasing and the level of social progress and economic development. How to break through the traditional industrial operation model has become the first issue of agricultural development.

Objective: Break the disadvantages of traditional breeding technology, scientific and efficient breeding help rural revitalization

Traditional breeding technology mainly relies on artificial statistics, regular medical examinations, and estimated weights for the weight of the column. The number of groups cannot be monitored in real time. Lack of advanced and appropriate scientific breeding systems. The team of the team uses Pose Former for multi -target key point detection and visual recognition. The target detection algorithm of the ductic classification is directly processed to the entire animal's image. Use the Point-Rend instance to divide the network and video understanding technology and other technologies to help poultry and animal-based breeding, realize the nutritional control, disease early warning and quality monitoring of poultry and herds, and then greatly improve the quality of poultry and livestock meat.

The picture shows some members of the social practice team to take a group photo

Practice: Visit in -depth on the spot, test technical results

The team reached related cooperation with the animal husbandry and water poultry breeding bases of Sichuan Agricultural University to obtain many data information of poultry and livestock. By sampled the samples of each single animal and labeled artificial data, it realized the related data sets of poultry and livestock.

The team applies artificial intelligence technology to the aquaculture of water poultry groups to achieve a healthy level of evaluation, scientifically records the growth indicators of the water poultry group, and associates with fleshy parameters. , Reduce the occurrence of large -scale diseases and improve economic benefits.

The picture shows the social practice team visiting the water poultry breeding base

"Sheep breeding, our current method is mainly to see if its hair is soft. If it is smooth and smooth, then it is normal and healthy. If its nose is moist, it means that it is healthy." Vaccine. "The staff who led the team to conduct animal examinations introduced to the inspectors that the breeding experience they shared was not only novel for the first members of the livestock breeding team, but also provided their practical knowledge and for the improvement of the construction of the big data smart detection system and experience.

The picture shows the team explore sheep breeding with "Yangcun Village Chief"

In response to the issues raised by the introducer, the team members focus on the application of the intelligent agricultural project in the health status monitoring of life and various diseases in livestock breeding. In the process of sheep growth and development, the use of target detection technology can identify sheep disease and help farmers find sheep disease earlier, which is of great significance to prevent the spread of the disease. Using gesture estimation and video understanding technology can help identify and predict the behavior of the sheep, warn the abnormal behavior, help farmers to better manage the herd, and promote the cultivation of farming, automation, and informatization. Continuous development.

Picture shows a list of sheep rings

The person in charge of the team introduced: "For the water poultry, we will rely on artificial intelligence to innovate the self -induced quantitative evaluation method of individual water poultry, forming a intelligent water poultry breeding ecological management system to promote the cultivation of farming, automation, and informatization. Continuous development of the direction. Facing animal husbandry, we tried to combine the record data of the flock growth process and the quality evaluation results of the sheep in all aspects of the sheep's quality evaluation results for data mining and scientific modeling, and then developed a cattle and sheep flock combining artificial intelligence and computer vision In the end, the smart health breeding system can eventually be used in a variety of cattle and sheep farming represented by cattle and sheep, and realizes the 'scientific', 'intelligent', and 'standardization' of cattle and sheep breeding.

Smart breeding helps rural industry reforms, and digital science empowers rural revitalization. Students have said that they will use the knowledge they have learned to contribute to the revitalization of the countryside, actively participate in the village, and perform their talents on fertile soil.

- END -