Target intelligent perception and identification system and method based on space coding

A technology of spatial coding and intelligent perception, applied in the field of target intelligent perception system based on spatial coding, can solve problems such as large amount of calculation data analysis, reduce equipment cost, avoid image post-processing process, and improve time efficiency.

Pending Publication Date: 2021-05-04
NORTHEASTERN UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although these help to achieve more accurate perception and recognition of targets, they also require a greater amount of calculations to complete complex data analysis and post-processing processes

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
  • Target intelligent perception and identification system and method based on space coding
  • Target intelligent perception and identification system and method based on space coding
  • Target intelligent perception and identification system and method based on space coding

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0021] In Embodiment 1, the target intelligent perception and recognition system based on spatial coding is used to recognize handwritten digits in the public data set MNIST. Among them, the MNIST handwritten digit data set used contains 60,000 training samples with labels and 10,000 verification samples with verification labels, each of which is a grayscale image composed of 28×28 pixels and the label It is one of the numbers from 0 to 9; use the one-to-many method to simplify the task of recognizing digits into ten binary classification tasks, that is, each time one class is used as the target group, and the remaining nine classes are used as the background group. The specific structure of the convolutional neural network for binary classification is as follows: figure 2 shown, and finally evaluate its performance using 5-fold cross-validation method. More specifically, before feeding the image data into the convolutional neural network, each image data is normalized, and ...

Embodiment 2

[0046] In this embodiment, the target intelligent perception and recognition system based on spatial coding is used to identify the direction of the target line segment. The specific implementation steps are as follows:

[0047] Step 1. Using the shape and direction features of the line segment to directly generate the weight value of each pixel unit in the space required for spatial encoding. The shape feature of the line segment determines that these weight values ​​are distributed in strips, and the direction feature of the line segment determines that these weight values ​​are distributed in different angles, so the weight values ​​of each pixel unit in the final generated space are distributed in strips at different angles, such as Figure 5 shown.

[0048] Step 2. Using the weight values ​​distributed in different angles in the strip shape generated in step 1 to control the spatial light modulator to perform spatial encoding, that is, to control the micromirror units in...

Embodiment 3

[0051] In this embodiment, the target intelligent perception and recognition system based on spatial coding is used to perform mean value filtering on the target, and the size of the mean value filter is 3×3 pixels. The specific implementation steps are as follows:

[0052] Step 1. Directly generate the weight value of each pixel unit in the space required for local space encoding according to the characteristics of the mean filter. Since the size of the mean filter is 3×3 pixels, the size of the final generated local space should be 3×3 pixels, and the weight value of each pixel unit in the local space is 1 / 9.

[0053] Step 2. Use the weight value of each pixel unit in the space generated in step 1 to control the spatial light modulator to perform spatial encoding, that is, control the pixel point and the micromirror unit at the corresponding position of the eight pixel points around the pixel point to be "on" state, the rest of the micromirror units are in the "off" state, ...

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 belongs to the field of optics, and discloses a target intelligent sensing and recognition system and method based on space coding, and the system comprises a coding module and an imaging module. The encoding module encodes the target recognition model into a weight value corresponding to each pixel unit in the space, and inputs the weight value into a spatial light modulator in the imaging module; and the imaging module performs spatial coding on the optical signal by using a spatial light modulator, acquires the spatially coded optical signal through a single-channel photoelectric detector, and realizes intelligent identification of a target by using a single-channel measurement value. According to the method, space coding is carried out on the collected optical signals, post-processing of the image is achieved while optical measurement is carried out, the complex image post-processing process is effectively avoided, and the time efficiency of target sensing and recognition is greatly improved. According to the method, the equipment cost caused by image data transmission, storage and the like can be reduced. Therefore, the method has important potential application value in the field of target perception and recognition.

Description

technical field [0001] The invention belongs to the field of optics, and relates to a target intelligent perception system and method based on spatial coding. Background technique [0002] Target perception and recognition is a comprehensive multidisciplinary application technology that detects fixed or moving targets and surrounding environment information (such as images, etc.) in a non-contact manner, and uses information processing technology to analyze and post-process the collected data , and then accurately identify the target, which has a wide range of applications in many fields such as industry, agriculture and national defense. In recent years, with the continuous development and improvement of target perception and recognition technology, the collected and extracted target feature information has become more and more abundant. Although these help to achieve more accurate perception and recognition of targets, they also require a greater amount of calculations to...

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): G06T9/00G06N3/08G06N3/04
CPCG06T9/002G06N3/08G06N3/045
Inventor 陈硕王东升张冯頔
Owner NORTHEASTERN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap