Goods deep learning training method based on machine vision and data sampling

Abstract

The invention discloses a goods deep learning training method based on machine vision and data sampling. The method comprises the following steps: S1, establishing an initial deep learning model of each goods; s2, the intelligent terminal collects basic data of the stored goods; s3, the service platform performs fuzzy judgment on the category of the goods; s4, performing user selection on all possible items of the goods; s5, the user selection result and the system judgment result are subjected to weighted calculation, and the final category of the goods is determined; s6, if system fuzzy judgment fails, feature point extraction is performed through the basic data, and a deep learning method is adopted to supplement and train the feature points into a corresponding goods initial deep learning model; and S7, accumulating the times of failure and success of system judgment. When the method is applied, a deep learning training method can be adopted to establish a classification and recognition model of goods stored in the intelligent terminal, and the classification and recognition model is continuously perfected through data sampling and feature supplementation.

Description

technical field [0001] The invention relates to the technical field of deep learning, in particular to a deep learning training method for goods based on machine vision and data sampling. Background technique [0002] With the rapid development of mobile Internet and smart terminal technology, various item storage terminals have appeared on the market, which can be used to store express items, waste and recycled items, household garbage, etc., but such item storage terminals usually only collect items After being stored, it is transported to the sorting center for manual sorting. The camera and X-ray scanning device installed on the smart terminal are only used for personnel monitoring and dangerous goods detection, and cannot realize the intelligent classification of the items stored in the smart terminal. In order to realize the intelligent classification of the items stored in the smart terminal, it is necessary to use the image recognition technology based on the recogni...

AI Technical Summary

Problems solved by technology

[0002] With the rapid development of mobile Internet and smart terminal technology, various item storage terminals have appeared on the market, which can be used to store express items, waste and recycled items, household garbage, etc., but such item storage terminals usually only collect items After storage, it is transported to the sorting center for manual sorting. The camera and X-ray scanning device installed on the smart terminal are only used for personnel monitoring and dangerous goods detection, and cannot realize the intelligent classification of the items stored in the smart terminal.

Images