Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Feature Quantity Calculation Device, Feature Quantity Calculation Method

A computing device and feature quantity technology, applied in the field of computer vision, can solve the problems of overlapping, increase in misjudgment, etc., and achieve the effect of high effect and high misjudgment

Inactive Publication Date: 2015-08-19
PANASONIC INTELLECTUAL PROPERTY CORP OF AMERICA
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0033] (1) There is the following problem: since the edge information is always extracted from the image, when there are many edges in the background pixels, the noise overlaps in the feature quantity, and the misjudgment increases

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
  • Feature Quantity Calculation Device, Feature Quantity Calculation Method
  • Feature Quantity Calculation Device, Feature Quantity Calculation Method
  • Feature Quantity Calculation Device, Feature Quantity Calculation Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach 1

[0085] First, various technical terms in this embodiment will be described.

[0086] "Pixel value" contains the brightness value.

[0087] "Edge strength" is information indicating the degree of change in pixel values. The edge strength is quantitatively represented by an "edge strength value" indicating the amount of change in the pixel value.

[0088] "Edge direction" means an edge gradient, and is the direction in which the edge strength changes. The edge direction is expressed quantitatively by the "edge direction value" which represents the direction in which the pixel value increases most in angle.

[0089] "Edge direction group" is a group of edge directions at a plurality of positions in which a predefined specific configuration relationship exists. The edge direction group is represented by a group of edge direction values ​​at each position.

[0090] The "correlation value" is information that quantitatively indicates the degree of correlation of edge intensities...

Embodiment

[0183] In this embodiment, how to define the dimensions of the histogram is arbitrary. Therefore, coHOG, LBP, etc., which are applicable and applicable to conventional methods, consider various feature quantities of two or more pixels. For comparison, non-patent document 1 (Dalai's HOG) was compared with scoHOG of a conventional method, and the effectiveness of this embodiment was verified.

[0184] Figure 15 It is a figure which shows the result of the accuracy effect verification using the feature quantity calculation apparatus 100 of this embodiment in the form of a table.

[0185] As the database used in the experiment, the INRIA dataset proposed in Non-Patent Document 1 and commonly used for algorithm evaluation of person detection is used. As images for learning, 2416 person images and 1218 background images without people are prepared. Regarding the background image, rectangular images of 10 parts randomly cut out of 1218 prepared images were used as background samp...

Embodiment approach 2

[0228] Embodiment 2 is an example of a case where a feature such as an LBP (Local Binary Pattern) feature is used instead of an edge direction value as a feature type for constructing a histogram feature.

[0229] Figure 17 It is a block diagram showing the configuration of a feature quantity calculation device according to Embodiment 2 of the present invention. right with Figure 5 The same constituent units are assigned the same reference numerals and descriptions of overlapping parts are omitted.

[0230] Such as Figure 17 As shown, the feature calculation device 200 includes a feature calculation unit 220 , a histogram creation unit 130 , and a histogram connection unit 140 .

[0231] The feature amount calculation unit 220 calculates an LBP feature from the input image data, and combines the above-mentioned LBP features of a plurality of pixels to calculate a feature amount between pixels.

[0232] Figure 18 is a block diagram showing the configuration of the feat...

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

A feature amount calculation apparatus calculates a feature amount of a target object from image data, and is provided with: a feature value calculator that calculates an edge direction and edge magnitude as input image data pixel-unit feature values; a feature amount calculator that has an edge direction group calculator that calculates a group of edge directions, and a correlation value calculator that takes all pixels or predetermined pixels among a plurality of pixels used in feature value calculation as pixels subject to correlation value calculation and calculates an edge magnitude correlation value between the pixels subject to correlation value calculation for each feature value; and a histogram creator that counts feature amounts in a histogram for each correlation value, and creates a histogram as a feature vector.

Description

technical field [0001] The present invention relates to a feature quantity calculation device, a feature quantity calculation method, and a program in the field of computer vision, particularly based on image detection and detection of the position and presence of a target object (hereinafter referred to as an object). Background technique [0002] Technology that detects the position of a person appearing in an image is expected to be effectively utilized in various applications, such as a video surveillance system, a driving support system for a car, or an automatic annotation system in an image or a video, etc. It conducts research. [0003] The scan window search type detection method is to perform a detailed raster scan on the input image with a variable-sized scan rectangular window, extract the image features in the scanned scan window, and use the offline (off line) method separately Based on the learned recognizer, it is judged whether there is an object object in ...

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 Patents(China)
IPC IPC(8): G06T7/60G06T7/00G06V10/50
CPCG06K2009/4666G06K9/4642G06K9/3241G06V10/255G06V10/50G06V10/467
Inventor 河合富美早田启介
Owner PANASONIC INTELLECTUAL PROPERTY CORP OF AMERICA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com