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

Depth image semantic segmentation method based on deep learning

A deep image and semantic segmentation technology, applied in neural learning methods, instruments, biological neural network models, etc., can solve problems such as difficulty in obtaining high segmentation accuracy

Active Publication Date: 2021-02-05
HANGZHOU NORMAL UNIVERSITY
View PDF0 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This has always been a difficulty in machine vision technology, and it has been difficult to obtain high segmentation accuracy before

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
  • Depth image semantic segmentation method based on deep learning
  • Depth image semantic segmentation method based on deep learning
  • Depth image semantic segmentation method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0016] The present invention will be further described below in conjunction with drawings and embodiments.

[0017] Such as figure 1 and 2 As shown, a depth map semantic segmentation method based on deep learning, specifically includes the following steps:

[0018] Step 1: Process the dataset and input the processed dataset into the ResNet network model.

[0019] The 1-1 dataset is mainly derived from the NYU-DepthV2 dataset, which consists of video sequences of various indoor scenes recorded by the RGB and Depth cameras of Microsoft Kinect. It has the following features: 1449 detailed labeled RGB and depth images; 464 different scenes from multiple cities; an instance number for each specific classification (e.g. bed 1, bed 2, bed 3, etc.)

[0020] 1-2 The data in the data set is preprocessed, redundant features are deleted, missing values ​​are processed, unreasonable data is removed, and features are normalized. Missing data were filled by a coloring scheme. Then put t...

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 depth image semantic segmentation method based on deep learning. The method comprises the following steps: 1, processing a data set and inputting the processed data set intoa ResNet network model; 2, inputting the data set processed in the step 1 into a ResNet network down-sampling stage, and fusing the information of the RGB image and the depth image in a down-samplingcoding stage by using a mode of respectively training and gradually fusing to obtain features extracted in the down-sampling coding stage; 3, inputting the features extracted in the ResNet network down-sampling coding stage in the step 2 into an up-sampling coding stage, adding an enhanced supervision module in the up-sampling coding stage process, optimizing a semantic segmentation result, and performing deeper feature extraction; and 4, performing depth image semantic segmentation by using the network model obtained by training in the step 3. According to the method, features are not lost ina deep network, and a reinforced supervision module is added in an up-sampling process to optimize a semantic segmentation result.

Description

technical field [0001] The invention relates to the technical field of semantic segmentation, in particular to a method for semantic segmentation of depth maps based on deep learning, which is used for identification of indoor environments. Background technique [0002] Semantic segmentation is one of the most important tasks in the field of computer vision. The understanding of the scene is the key technology for the robot to realize artificial intelligence and interact with the outside world. In order for a robot to assist humans, or even complete work independently, it must have two basic capabilities, one is to be able to recognize what types of objects exist in the working environment, and the other is to recognize the location of specific types of objects. [0003] At present, the goal of semantic segmentation technology researched by scholars at home and abroad is to simultaneously complete the two tasks of object classification and object detection through a single ...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/34G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V10/267G06N3/045G06F18/25
Inventor 盛伟国陈浩天
Owner HANGZHOU NORMAL UNIVERSITY