The layman does not know that the most difficult thing to grow cucumber is not planting and harvesting, because these jobs are relatively standardized. The most difficult thing is the sorting of cucumbers. After the cucumbers are mature, they are of different sizes and maturity. How to sort them out according to the grades used to be manual, time-consuming and labor-consuming, and the cost is relatively large. In the high season of cucumber harvest, farmers spend more than 8 hours a day sorting cucumbers. Now an artificial intelligence system based on visual recognition developed by tensorflow system can easily complete this work. Specifically, it is to take pictures of different cucumbers, so that artificial intelligence can learn what kind of cucumbers should be divided into which level, and then do automatic sorting on the assembly line, which greatly improves the efficiency of sorting and makes it easier for farmers to operate their farms.

Although all kinds of soil, climate sensors and remote sensing images have come out one after another, the application in agriculture is far from enough, and the optimization of crop production is still limited, especially in the prediction of late blight, whitefly and other pests. With the help of artificial intelligence analysis, the farm can deal with infected leaves in time and reduce the loss as soon as possible. In the past, the detection of diseases and insect pests required manual inspection, once found not in time, will lead to the death of large crops. Manual inspection is time-consuming and laborious, and there may be omissions. The introduction of artificial intelligence can provide continuous monitoring and forecasting, and reduce the loss caused by diseases and insect pests.

In the past, we saw a plant we didn't know in the wild, we had to consult the data to know what the flowers and plants were, but now we can take photos of the flowers and plants through various map recognition software and scan them to know, which is computer image recognition technology. Nowadays, with the help of machine learning and deep learning, the accuracy of intelligent image recognition is getting higher and higher, and the application is far more than these.

In the agricultural field, artificial intelligence can realize the functions of soil detection, pest protection, yield prediction, and early warning of livestock and poultry diseases. In the field of soil detection, intelinair has developed a UAV, which can take soil photos through similar NMR technology, and determine soil fertility and accurately determine the crops suitable for planting through computer intelligent analysis.

Intelligent wearable products are mainly used in animal husbandry, which can collect the individual information of livestock and poultry in real time, identify the health status of livestock and poultry, predict the estrus period, and detect the feeding status through machine learning technology, so as to obtain the corresponding disposal in time. If the cow is sick, the information will be automatically processed through the analysis software.