Moving object detection method and system

Patent metadata

Field	Value
Publication number	US-2018365845-A1
Application number	US-201615737155-A
Country	US
Kind code	A1
Filing date	Dec 7, 2016
Priority date	Aug 19, 2016
Publication date	Dec 20, 2018
Grant date	—

Abstract

A moving object detection method and a moving object detection system are provided. The method includes: predetermining a background image corresponding to a scene monitored by a video monitoring device; performing a subtraction processing on a grayscale image to be detected and the background image to acquire a difference image; and binarizing the difference image and determining a moving object in the grayscale image to be detected, where the determining the background image includes: dividing a first grayscale image frame and a second grayscale image frame in a grayscale image frame sequence captured by the video monitoring device into image blocks to acquire a first image block set and a second image block set respectively, and determining the background image using a difference between the first image block set and the second image block set.

First claim

1 . A moving object detection method, comprising:
predetermining a background image corresponding to a scene monitored by a video monitoring device; performing a subtraction processing on a grayscale image to be detected and the background image to acquire a difference image; and binarizing the difference image and determining a moving object in the grayscale image to be detected, wherein the determining the background image comprises:
dividing a first grayscale image frame and a second grayscale image frame in a grayscale image frame sequence captured by the video monitoring device into image blocks to acquire a first image block set and a second image block set respectively, and
determining the background image using a difference between the first image block set and the second image block set.

2 . The moving object detection method according to claim 1 , wherein a time instant at which the grayscale image to be detected is captured is greater than or equal to a time instant at which the first grayscale image frame is captured, the time instant at which the first grayscale image frame is captured is greater than a time instant at which the second grayscale image frame is captured, and there are N grayscale image frames between the first grayscale image frame and the second grayscale image frame, wherein N is a positive integer.

3 . The moving object detection method according to claim 2 , wherein the dividing the first grayscale image frame and the second grayscale image frame in the grayscale image frame sequence captured by the video monitoring device into image blocks comprises:
dividing the first grayscale image frame into K image blocks which do not overlap with each other and include all pixels in the first grayscale image frame, to acquire the first image block set, and dividing the second grayscale image frame into K image blocks which do not overlap with each other and include all pixels in the second grayscale image frame, to acquire the second image block set, wherein K is a positive integer, the image blocks have the same size, and the K image blocks in the first image block set have a one-to-one correspondence with the K image blocks in the second image block set.

4 . The moving object detection method according to claim 3 , wherein the determining the background image using the difference between the first image block set and the second image block set comprises:
calculating the difference between the first image block set and the second image block set; and acquiring the background image using the difference and the first image block set.

5 . The moving object detection method according to claim 4 , wherein the calculating the difference between the first image block set and the second image block set comprises:
calculating an image block grayscale difference between each image block in the first image block set and a corresponding image block in the second image block set, to acquire an image block grayscale difference set, wherein an i-th element in the image block grayscale difference set is calculated according to an equation:
d i =|g t,i −g t−N,i |,
wherein g t,i represents an i-th image block in the first image block set, which is represented by g t , g t−N,i represents an i-th image block in the second image block set, which is represented by g t−N , and d i represents an image block grayscale difference between the i-th image block in the first image block set and the i-th image block in the second image block set, wherein i=1, 2, . . . , K; and calculating a difference between each image block in the first image block set and a corresponding image block in the second image block set using the image block grayscale difference set to acquire a difference set, wherein an i-th element in the difference set is calculated according to an equation:

s
i

=

∑

1
≤
x
≤
n

,

1
≤
y
≤
m





d
i



(

x
,
y

)

,

wherein n represents the length of each image block, m represents the width of each image block, both n and m are in units of pixels, d i (x,y) represents a grayscale difference in d i corresponding to an (x,y)-th pixel in the image block, and s i represents a difference between an i-th image block in the first image block set and an i-th image block in the second image block set.

6 . The moving object detection method according to claim 5 , wherein the acquiring the background image using the difference and the first image block set comprises:
converting the difference set to a determination set using a preset conversion equation:

w
i

=

{

1
,

s
i

≤

TH
1

0
,

else

,

wherein TH 1 represents a preset difference threshold, and w i represents an i-th determination element in the determination set; and
inputting the determination set and the first image block set into a preset first background image construction equation to acquire the background image, wherein the first background image construction equation is expressed as:

b
i

=

{

g

t
,
i

Assignees

• Univ Soochow

Inventors

• Zhang Li
• Pei Tingwei
• Wang Bangjun
• Zhang Zhao
• Li Fanzhang

Classifications

• G06T2207/30221

  Sports video; Sports image · CPC title

• G06T2207/20021

  Dividing image into blocks, subimages or windows · CPC title

• G06T7/254 Primary

  involving subtraction of images · CPC title

• G06T2207/10016

  Video; Image sequence · CPC title

• G06T7/231

  using full search · CPC title