Unlock instant, AI-driven research and patent intelligence for your innovation.

Parcel counting method and system based on deep learning and multi-target tracking technology

A technology of multi-target tracking and deep learning, which is applied in the field of package counting methods and systems based on deep learning and multi-target tracking technology, can solve the problems of easy deviation and low efficiency of piece counting, and achieve the effect of solving low efficiency of piece counting

Pending Publication Date: 2020-04-10
SUN YAT SEN UNIV
View PDF11 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In order to solve the problem that the existing logistics package counting method is prone to deviation and low piece counting efficiency, the present invention provides a package counting method and system based on deep learning and multi-target tracking technology

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Parcel counting method and system based on deep learning and multi-target tracking technology
  • Parcel counting method and system based on deep learning and multi-target tracking technology

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0037] A package counting method based on deep learning and multi-target tracking technology, such as figure 1 shown, including the following steps:

[0038] S1. Obtain the logistics package data and perform preprocessing; wherein the logistics package data is divided into a training set and a verification set;

[0039] In this step, the logistics package data is obtained by setting an industrial camera on the package conveyor belt and obtaining continuous package pictures according to the preset frequency, so as to obtain all logistics package data within a period of time. In order to avoid missing logistics packages, use The front and back frames are continuous in content, so that the packages in various positions in the camera's field of view can also be obtained, thereby avoiding uneven data distribution, and allowing the model to pay attention to various positions of the picture during subsequent model training;

[0040] A total of 10,000 parcel images were obtained, of ...

Embodiment 2

[0053] The present embodiment 2 is the package counting system based on deep learning and multi-target tracking technology proposed based on the package counting method based on deep learning and multi-target tracking technology in embodiment 1, such as figure 2 shown, including:

[0054] The data acquisition and preprocessing module 1 is used to obtain and preprocess the parcel data of the logistics; wherein the parcel data of the logistics is divided into a training set and a verification set;

[0055] The package detection model training module 2 is used to train the preset package detection model based on deep learning through the training set, and test and parameterize the trained package detection model through the verification set; obtain the final package detection model;

[0056] Parcel detection module 3, for obtaining real-time logistics parcel video and carrying out key frame sampling to it, input described final parcel detection model to detect thereby obtain th...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a parcel counting method and a system based on deep learning and a multi-target tracking technology. The parcel counting method comprises the following steps: firstly, obtaining parcel data of logistics and preprocessing the parcel data; training a preset parcel detection model based on deep learning through the training set, and performing testing and parameter adjustmenton the trained parcel detection model through the verification set; obtaining a final package detection model; acquiring a real-time logistics package video, performing key frame sampling on the logistics package video, and inputting the logistics package video into the final package detection model for detection so as to acquire package position information of each key frame; and counting the packages by adopting a multi-target tracking algorithm according to the package position information of the key frame. Automatic counting of logistics packages is carried out on the basis of a modern computer vision technology and multi-target tracking counting, logistics package vision and position information is mined, and even if package pictures obtained in real time contain densely and irregularly-placed packages, the packages can be accurately counted.

Description

technical field [0001] The invention relates to the technical field of package counting, in particular to a package counting method and system based on deep learning and multi-target tracking technology. Background technique [0002] In recent years, with the vigorous development of e-commerce and various shopping festival promotions, the order volume of the express delivery logistics industry has also increased like a tsunami, which has also brought many challenges to the traditional express delivery industry, including package statistics, inbound and outbound checks, and other issues. At present, the express logistics industry generally adopts the manual piece counting method, but the manual piece counting method is intensive and low in efficiency; there are also some package counting systems based on sensor technology, but when the number of packages is large or dense, since the sensor technology relies on distance, often resulting in missed inspections, resulting in devi...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06T7/00G06T7/246
CPCG06T7/0002G06T7/246G06T2207/10016G06T2207/10024G06T2207/20081G06T2207/20084G06T2207/30242
Inventor 严晓威吴发明马锦华李沁航
Owner SUN YAT SEN UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com