Agricultural picked object recognition algorithm based on database

Abstract

The invention belongs to the technical field of image recognition, and particularly relates to an agricultural picked object recognition algorithm based on a database, which is characterized in that a computer sets a maturity index according to a plurality of indexes such as colors, sizes and wrinkles of fruits to be picked and characteristics associated with maturity height, and establishes a model for the maturity degree and storage time of the fruits; the computer carries out artificial intelligence recognition on fruit pictures collected by the picking robot and transmits fruit types and coordinate information to a motion control unit of the picking robot, and finally grabbing, sorting and obstacle-avoiding walking are achieved. Compared with the prior art, the method has the beneficial effects that the maturity of strawberries can be divided into multiple grades according to market requirements, the maturity of picked strawberries with long storage time or long transportation distance is slightly lower, so that the distance of mailing and selling areas can be selected according to the maturity, the loss of the strawberries in transportation is reduced, and the economic benefit is increased. Sales areas are expanded, and economic benefits are improved.

Description

technical field [0001] The invention belongs to the technical field of image recognition, and in particular relates to a recognition algorithm of agricultural picking objects based on a database. Background technique [0002] There are many precautions for fruit picking, and the picking steps are cumbersome. However, in many areas of our country, fruit picking is still mainly done manually, and production mechanization and modelization have not been realized to a large extent. Simply relying on labor is not only inefficient, but also the quality of finished fruit products cannot be guaranteed. This picking method is time-consuming and labor-intensive, and requires the cooperation of multiple people to complete. Most of the picking machines in the prior art cannot be sorted and packed according to the maturity and size of the fruit during the picking process. They are usually picked in the orchard and transported to the sorting site by car, and then sorted manually or mechan...

AI Technical Summary

Problems solved by technology

At this time, according to the first-in-first-out warehouse management, it will undoubtedly affect the user's experience of tasting fruit

[0004] In modern three-dimensional strawberry greenhouses, strawberries have slight color differences and small size differences. Therefore, it is necessary to develop an artificial intelligence fruit state recognition and detection system that can detect and judge the maturity of fruits during the picking process. The time difference between the refrigeration and transportation process assists the warehouse staff in the operation of entering and leaving the warehouse, and improves the precision and accuracy of the management of the warehouse entering and leaving the warehouse. At present, there are few reports in the relevant literature

Images