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

Method and system for detecting a human in a test image of a scene acquired by a camera

A technology for testing images and cameras, applied in the field of computer vision, can solve problems such as inappropriate real-time application, and achieve the effect of fast and accurate person detection

Inactive Publication Date: 2009-01-28
MITSUBISHI ELECTRIC CORP
View PDF2 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0016] Furthermore, the Dalal & Triggs method is only able to process a 320×240 pixel image at approximately one frame per second, and even a very sparse scanning method can only compute approximately 800 detection windows per image.
Therefore, the Dalal & Triggs method is inappropriate for real-time applications

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
  • Method and system for detecting a human in a test image of a scene acquired by a camera
  • Method and system for detecting a human in a test image of a scene acquired by a camera

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] FIG. 1 is a block diagram of a system and method for training ( 10 ) a classifier 15 using a set 1 of training images and detecting ( 20 ) a person 21 in one or more test images 101 using the trained classifier 15 . The method used to extract features from training images and test images is the same. Since the training is performed in one preprocessing stage, the training will be described later.

[0027] Figure 2 shows a method 100 for detecting a person 21 in one or more test images 101 of a scene 103 captured by a camera 104 according to an embodiment of our invention.

[0028] First, we determine (110) the gradient for each pixel. For each cell, we determine a weighted sum of the gradient orientations of the pixels in that cell, where the weights are based on the magnitude of the gradient. These gradients are sorted into 9 bars of the Histogram of Gradients (HoG) 111. We store (120) the integral image 121 for each strip in the HoG in memory. This results in 9 in...

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 method and system is presented for detecting humans in images of a scene acquired by a camera. Gradients of pixels in the image are determined and sorted into bins of a histogram. An integral image is stored for each bin of the histogram. Features are extracted fom the integral images, the extracted features corresponding to a subset of a substantially larger set of variably sized and randomly selected blocks of pixels in the test image. The features are applied to a cascaded classifier to determine whether the test image includes a human or not.

Description

technical field [0001] The present invention relates generally to computer vision, and more particularly to detecting people in images of scenes captured by cameras. Background technique [0002] It is relatively easy to detect human faces in a sequence of images of a scene captured by a camera. However, detecting people remains a challenge because their appearance varies greatly due to clothing, joints, and lighting conditions in the scene. [0003] There are mainly two types of methods that use computer vision methods to detect people, see D.M.Gavrila, "The visual analysis of human movement: A survey", Journal of Computer Vision and Image Understanding (CVIU), vol.73, no.1, pp. 82-98, 1999. One class of methods uses section-based analysis, while the other uses single detection window analysis. Different features and different classifiers for these methods are known. [0004] Part-based approaches aim to deal with the large variability in human appearance due to human j...

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): G06K9/00G06K9/46
CPCG06K9/00369G06K9/4614G06K9/4647G06V40/103G06V10/446G06V10/507
Inventor 什穆埃尔·阿维丹朱强
Owner MITSUBISHI ELECTRIC CORP