Machine control method and system based on object recognition

A method includes acquiring one or more images of a set of unorganized articles within a first machine; determining one or more article types of the unorganized article set from the one or more images, including: dividing a corresponding image in the one or more images into a plurality of correspond...

Full description

Saved in:
Bibliographic Details
Main Authors OU ZHICAI, TIAN YUNKE, HE QING
Format Patent
LanguageChinese
English
Published 10.05.2022
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:A method includes acquiring one or more images of a set of unorganized articles within a first machine; determining one or more article types of the unorganized article set from the one or more images, including: dividing a corresponding image in the one or more images into a plurality of corresponding sub-regions; feature detection is carried out on the multiple corresponding sub-regions to obtain multiple corresponding region feature vectors, and the region feature vectors of the sub-regions indicate characteristics of multiple predefined local article features of the sub-regions; generating an integrated feature vector by combining the corresponding plurality of regional feature vectors; applying a plurality of binary classifiers to the integrated feature vector; and selecting machine settings for the first machine based on one or more item types determined in the set of unorganized items. 一种方法包括:采集第一机器内的无组织物品集合的一个或多个图像;从一个或多个图像中确定无组织物品集合的一个或多个物品类型,包括:将一个或多个图像中的相应图像划分为相应的多个子区域;对所述相应的多个子区域进行特征检测,以获取相应的多个区域特
Bibliography:Application Number: CN202080065025