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

Method of detecting moving objects

A technology for moving objects and images, applied in image analysis, image enhancement, instruments, etc., can solve problems such as difficult to recreate standard backgrounds

Pending Publication Date: 2022-01-28
THE EDGE CO SRL
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the case of mobile devices, both the background and the foreground are moving, so it is difficult if not impossible to recreate the standard background for detection, especially if the objects are very small

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 of detecting moving objects
  • Method of detecting moving objects
  • Method of detecting moving objects

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0025] figure 1 A diagram of a preferred embodiment of the method according to the invention is shown.

[0026] The method of detecting moving objects requires making a film (or video) of the space to be observed.

[0027] According to the invention, the detection method comprises a step 1: extracting at least one first image (f1) and at least one second image (f2) from the film.

[0028] The images (frames) that make up a video are in color and made up of pixels. Specifically, the resolution is 1280×720 pixels, with a variable frame rate between 24 and 60 frames per second.

[0029] Preferably, the second image (f2) is continuous with the first image (f1).

[0030] In step 2, corner points are detected in the first image (f1) and in the second image (f2). Specifically, in step 2, a series of edges and corners are detected in the first image (f1) and in the second image (f2). A corner is the intersection of two or more edges.

[0031] Advantageously, the points are detec...

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 of detecting moving objects, comprising the following steps: extract at least one first image (fl) and at least one second image (f2) from a video of an observed space; detect comer points in the first image (fl) and in the second image (f2); using the points identified, extract at least one feature vector; identify corresponding points in the first image (fl) and in the second image (f2); determine the homography to perform image registration between the first image (fl) and the second image (f2); apply a warping of the second image (f2) based on the homography to obtain a warped second image (f2_w); calculate the difference in individual pixels between the warped second image (F2_w) and the first image (fl) to identify the pixels that differ; obtain a third image (diff) by comparing the second warped image (F2_w) against the first image (fl); identify a bounding box (bb) in the third image (diff) for each group of adjacent pixels; calculate the optical flow based on the first image (fl) and the warped second image (f2w); extract from the optical flow a matrix (mag) corresponding to the magnitude of the optical flow; for each bounding box (bb) drawn on the third image (diff), check whether the optical flow magnitude of a detected pixel group is greater than the average magnitude of the surrounding pixels; discard the bounding box (bb) if the optical flow magnitude is not greater than the average magnitude of the surrounding pixels; use a blob detector for each bounding box (bb) drawn on the third image (diff); discard the bounding box (bb) if there are no blobs; and assign the detection of an object to the undiscarded bounding boxes (bb).

Description

technical field [0001] The invention relates to a method for detecting a moving object. Background technique [0002] Various methods are currently known to detect moving objects using "motion detection" algorithms. These algorithms are based on computer vision techniques, using video captured by the camera as their starting point. [0003] Cameras are usually stationary. Their frames can vary, but they cannot move close to objects. To move closer or follow moving objects, drones, that is to say small aircraft that are radio-controlled from the ground, can be used. [0004] Today, drones have various applications: security and tracking, environmental and building surveillance, remote sensing, collection of qualitative and quantitative data on a specific territory after analyzing emitted or reflected electromagnetic radiation, and video recording. The video footage specifically enables the detection of objects moving in the vicinity of the drone. [0005] An interesting ...

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): G06T7/246G06T7/269
CPCG06T7/246G06T7/269G06T2207/10016G06T2207/20164G06T7/215G06T7/12G06T7/73
Inventor F·马希
Owner THE EDGE CO SRL