CN103778629A - Background model real-time updating method for non-coherent radar image - Google Patents
Background model real-time updating method for non-coherent radar image Download PDFInfo
- Publication number
- CN103778629A CN103778629A CN201410017422.XA CN201410017422A CN103778629A CN 103778629 A CN103778629 A CN 103778629A CN 201410017422 A CN201410017422 A CN 201410017422A CN 103778629 A CN103778629 A CN 103778629A
- Authority
- CN
- China
- Prior art keywords
- background model
- pixel
- sample
- new samples
- radar image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a background model real-time updating method for a non-coherent radar image and the method is used for low-altitude airspace security monitoring. In the method, a background model sample set is established for each pixel in a radar image and in a background model updating process, the value of each pixel in a current frame of image is compared with the background model sample set of the pixel of a current background model are compared and if the pixel values belong to a sample of the background model, then the pixel values replace randomly a specific sample in the current background model. At the same time, the pixel values are used to update randomly a background model sample set of a specific adjacent-domain pixel. The update cycle of the background model also adopts a random setting method. The background model real-time updating method for the non-coherent radar image prolongs the time that each sample persists in a background model set under a condition that the number of background model samples is kept unchanged so that time coverage range of the background model samples is expanded.
Description
Technical field
The present invention relates to a kind of background model real time updating method of non-coherent radar image, belong to low altitude airspace security monitoring technical field, relate to radar image and process and target detection.
Background technology
Once non-coherent radar have cost low, set up conveniently, the feature such as working alone property is strong, be the important means of spatial domain security monitoring.Non-coherent radar itself does not possess the function that moving-target detects, ripe radar surveillance system adopts image pick-up card that scope indicating image is transferred to computing machine conventionally, by rear end, the algorithm of target detection based on image is processed it again, therefrom extracts moving-target information.Background differential technique is the most frequently used moving target detection technique.But, because the region of system monitoring is low altitude airspace, background environment complexity, noise is strong, and background object glint is larger, and has certain random character.Therefore, set up background model accurately, and adopt suitable method to carry out real-time update to it, become the key that improves target detection ability.
Traditional background model update method, conventionally utilize the new background sample of pixels of extracting in current frame image to remove to replace the longest sample of retention time in original background model, the time of causing each sample to retain in background model is identical, has limited the time coverage of background model sample.
Summary of the invention
The object of the present invention is to provide a kind of background model real time updating method of non-coherent radar image, the method is applicable to detect based on the moving-target of non-coherent radar image, has widened the time coverage of background model sample.
The background model real time updating method of non-coherent radar image of the present invention, comprises the steps:
Step 1, for the each pixel in radar image is set up background model;
In radar image, gray-scale value corresponding to each pixel (x, y) represented by v (x, y), uses v
irepresent i background model sample, the background model M (x, y) of each pixel (x, y) is expressed as by N the background model sample set extracting in N two field picture before:
M(x,y)＝{v
1,v
2,...,v
N} (1)
N background model sample in formula (1) set, initially obtains: for i (1≤i≤N) two field picture in radar image sequence, be communicated with neighborhood N from 8 of pixel (x, y) by the following method
gin (x, y), extract i background model sample v
i:
Wherein, coordinate
at neighborhood N
gin (x, y), select at random,
denotation coordination
the gray-scale value at place.
Step 2, reads a frame radar image, to the new samples classification of each pixel;
According to the background model M (x, y) of pixel (x, y), the new samples v (x, y) of pixel in current frame image is classified, specifically: centered by v (x, y), R is radius, set up spherical territory S
r(v (x, y)), if the sample number in the common factor of the background model sample set of this spherical territory and M (x, y) is not less than threshold value #
min, new samples v (x, y) is judged as to background model sample, otherwise new samples is judged as foreground target.The common factor of spherical territory and background model sample set is expressed as:
#{S
R(v(x,y))∩{v
1,v
2,...,v
N}} (3)
Step 3, for the new samples that is judged to be background model, upgrades the background model of respective pixel;
For the new samples v (x, y) that is judged to be background model, between generation 0～1, random decimal θ, if θ is greater than threshold value T, upgrades pixel (x, y) background model, otherwise does not upgrade.
In the time carrying out background model renewal, generate the random integers i between 0～N, and replace i sample v in background model with new samples v (x, y)
i.
Step 4, for the new samples that is judged to be background model, upgrades neighborhood territory pixel sample;
For the new samples v (x, y) that is judged to be background model, random decimal θ between generation 0～1, if θ is greater than threshold value T, is communicated with pixel background model in neighborhoods to 8 of pixel (x, y) and upgrades, otherwise do not upgrade.
When carrying out in neighborhood pixel background model while upgrading, select at random the pixel coordinate (x in 8 connection neighborhoods
nG, y
nG), generate the random integers i between 0～N, with new samples v (x, y) replacement pixel (x
nG, y
nG) background model in i sample v
i.
The background model real time updating method of non-coherent radar image provided by the invention, compared with prior art, in the situation that keeping background model sample size constant, extend the time that each sample retains in background model set, widen the time coverage of background model sample.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of the background model real time updating method of non-coherent radar image of the present invention;
Fig. 2 is the original non-coherent radar image of a frame and certain location of pixels of the embodiment of the present invention.
Embodiment
Background model real time updating method the present invention being proposed below in conjunction with the result of certain non-coherent radar image sequence in accompanying drawing illustrates and describes.
In non-coherent radar background, comprise a large amount of stationary objects, wherein major part belongs to non-rigid targets (woods, meadow, the water surface etc.), echo strength big rise and fall, and background edge noise jamming is strong, detects and causes difficulty to low altitude small target.The inventive method is set the background model update cycle at random, and replace at random the sample in background model, extended the retention time of each sample in background model, as shown in Figure 1, the background model real time updating method of non-coherent radar image of the present invention comprises that concrete steps are as follows to its flow process:
Step 1, background model initializing; For the each pixel in radar image is set up background model.
First, in radar image, gray-scale value corresponding to each pixel (x, y) represented by v (x, y), and demarcates each background model sample v with index i
i, i is integer and 1≤i≤N, the background model M (x, y) of each pixel (x, y) forms background model sample set by N the background model sample extracting in multiple image before:
M(x,y)＝{v
1,v
2,...,v
N} (1)
For the first two field picture in radar image sequence (time t=0), can be communicated with neighborhood N from 8 of pixel (x, y)
gin (x, y), extract background model sample, be expressed as M
0(x, y):
Coordinate in above formula
can be at neighborhood N
gin (x, y), select at random the gray-scale value of certain pixel
may repeatedly be selected, also may be never selected.V
1be initially M
0(x, y).
Continue to read N-1 frame radar image, in every two field picture, identical with method shown in formula (2), obtain the background model sample value of pixel (x, y) in every two field picture.For i (1≤i≤N) two field picture in radar image sequence, be communicated with neighborhood N from 8 of pixel (x, y)
gin (x, y), extract i background model sample v
i:
finally obtain the background model suc as formula the pixel (x, y) shown in (1).
As shown in Figure 2, image size is 456 × 456 to the original non-coherent radar image of one frame, and true origin is in the image upper left corner, and to the right, vertically downward, in figure, the pixel coordinate of " * " position is (138,360) to Y-axis to X-axis level.The background model sample set of this pixel amounts to and comprises 10 samples, i.e. N=10, and the gray-scale value that each sample is pixel, background model is as follows:
M(138,360)＝{10,10,12,24,22,11,12,34,12,26} (3)
Step 2, reads a frame radar image, to the new samples classification of each pixel.
According to the current background model M (x, y) of pixel (x, y), the new samples v (x, y) of pixel (x, y) in current frame image is classified, judge whether new samples belongs to background model sample set.Centered by v (x, y), R is radius, sets up spherical territory S
r(v (x, y)), R is integer.If the sample number in the common factor of this spherical territory and background model sample set is not less than threshold value #
min, the new samples v (x, y) of this pixel is judged as to background model sample, proceed step 3 and 4; Otherwise the new samples v (x, y) that judges this pixel is not background model sample, be foreground target, do not carry out step 3 and 4 below.The common factor of spherical territory and background model sample set is expressed as:
#{S
R(v(x,y))∩{v
1,v
2,...,v
N}} (4)
In the embodiment of the present invention, the grey scale pixel value of establishing current frame image " * " position coordinate is v (138,360)=15, and centered by this new samples, R=5 is radius, sets up spherical territory S
r=5(v (138,360)), threshold value #
min=5, the common factor of this spherical territory and background model sample set is
#{S
R＝5(v(138,360))∩M(138,360)}＝{10,10,12,11,12,12} (5)
In occuring simultaneously, have 6 samples, be greater than threshold value #
minso this sample is background sample.
Step 3, for the new samples that is judged to be background model, upgrades the background model of respective pixel.In renewal, realize update cycle and the more random setting of new samples.
For the new samples v (x, y) that is judged to be background model, determine whether to use it for background model renewal by generating the method for random decimal θ between 0～1.If θ is greater than threshold value T, pixel (x, y) background model is upgraded, otherwise do not upgraded.Threshold value T is set by the user, and is the data between 0～1.
In the time carrying out background model renewal, adopt equally random device to select more new samples.Generate the random integers i between 0～N, and replace i sample v in background model with v (x, y)
i.
In the embodiment of the present invention, establish threshold value T=0.5, the random decimal of generation is θ=0.68, background model is upgraded, and replaces certain sample in original background model with new samples v (138,360)=15.Generate random integers i=6, replace the 6th sample of background model, the background model after renewal is
M(138,360)＝{10,10,12,24,22,15,12,34,12,26} (6)
Step 4, for the new samples that is judged to be background model, upgrades neighborhood territory pixel sample.
For the new samples that is judged to be background model, can be used for equally upgrading the background model that 8 of pixel (x, y) is communicated with neighborhood.
The setting of update cycle is identical with step 3, adopts the method that generates random decimal between 0～1 to judge whether to upgrade.In the embodiment of the present invention, the random decimal of establishing generation is θ=0.52, is greater than threshold value T=0.5, upgrades neighborhood territory pixel background model with new samples v (138,360)=15.
The random method of same employing selects to need the neighborhood territory pixel position and replacement sample of renewal.If (x
nG, y
nG) be the pixel coordinate in 8 connection neighborhoods, comprise (x-1, y-1), (x-1, y), (x-1, y+1), (x, y-1), (x, y+1), (x+1, y-1), (x+1, y) and (x+1, y+1).Random definite neighborhood coordinate (x
nG, y
nG) afterwards, identical with step 3, generate the random integers i between 0～N, and with v (x, y) replacement pixel (x
nG, y
nG) background model in i sample v
i.
In the embodiment of the present invention, establish the neighborhood territory pixel coordinate (x that random generation need be upgraded
nG, y
nG)=(138,361), the original background model of this pixel is:
M(138,361)＝{15,12,14,26,26,17,22,28,22,28} (7)
Generate random integers i=9, replace the 9th sample of this pixel background model, the background model after renewal is
M(138,361)＝{15,12,14,26,26,17,22,28,15,28} (8)
Utilizing present frame radar image to carry out after background model renewal, can continue to read next frame radar image, then go to step 2 continuation and carry out.
Claims (1)
1. a background model real time updating method for non-coherent radar image, is characterized in that, comprises the steps:
Step 1, for the each pixel in radar image is set up background model;
In radar image, gray-scale value corresponding to each pixel (x, y) represented by v (x, y), uses v
irepresent i background model sample, the background model M (x, y) of each pixel (x, y) is expressed as by N the background model sample set extracting in N two field picture before:
M(x,y)＝{v
1,v
2,...,v
N} (1)
N background model sample in formula (1) set, initially obtains: for i (1≤i≤N) two field picture in radar image sequence, be communicated with neighborhood N from 8 of pixel (x, y) by the following method
gin (x, y), extract i background model sample v
i:
Wherein, coordinate
at neighborhood N
gin (x, y), select at random,
denotation coordination
the gray-scale value at place;
Step 2, reads a frame radar image, to the new samples classification of each pixel;
For pixel (x, y), according to corresponding background model M (x, y), the new samples v (x, y) of pixel in current frame image is classified, specifically: centered by v (x, y), R is radius, set up spherical territory S
r(v (x, y)), if the sample number in the common factor of the background model sample set of this spherical territory and M (x, y) is not less than threshold value #
min, new samples v (x, y) is judged as to background model sample, otherwise new samples is foreground target;
Step 3, for the new samples that is judged to be background model, upgrades the background model of respective pixel;
For the new samples v (x, y) that is judged to be background model, between generation 0～1, random decimal θ, if θ is greater than threshold value T, upgrades pixel (x, y) background model, otherwise does not upgrade;
In the time carrying out background model renewal, generate the random integers i between 0～N, and replace i sample v in current background model with new samples v (x, y)
i;
Step 4, for the new samples that is judged to be background model, upgrades the background model sample of pixel in neighborhood;
For the new samples v (x, y) that is judged to be background model, random decimal θ between generation 0～1, if θ is greater than threshold value T, is communicated with pixel background model in neighborhoods to 8 of pixel (x, y) and upgrades, otherwise do not upgrade;
When carrying out in neighborhood pixel background model while upgrading, select at random the pixel coordinate (x in 8 connection neighborhoods
nG, y
nG), generate the random integers i between 0～N, with new samples v (x, y) replacement pixel (x
nG, y
nG) background model in i sample v
i.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410017422.XA CN103778629A (en) | 2014-01-15 | 2014-01-15 | Background model real-time updating method for non-coherent radar image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410017422.XA CN103778629A (en) | 2014-01-15 | 2014-01-15 | Background model real-time updating method for non-coherent radar image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103778629A true CN103778629A (en) | 2014-05-07 |
Family
ID=50570826
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410017422.XA Pending CN103778629A (en) | 2014-01-15 | 2014-01-15 | Background model real-time updating method for non-coherent radar image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103778629A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105184814A (en) * | 2015-07-27 | 2015-12-23 | 成都天奥信息科技有限公司 | Moving target detecting and tracking method based on multi-frame radar image |
| CN106023259A (en) * | 2016-05-26 | 2016-10-12 | 史方 | Method and device for detecting moving target frequency |
| CN108491796A (en) * | 2018-03-22 | 2018-09-04 | 电子科技大学 | A kind of time domain period point target detecting method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102314691A (en) * | 2011-06-30 | 2012-01-11 | 北京平安视讯科技有限公司 | Background model based on multiple information integration |
| CN102629383A (en) * | 2012-02-28 | 2012-08-08 | 湖南大学 | Motion object detection method based on random strategy |
| CN102665068A (en) * | 2012-04-26 | 2012-09-12 | 中南林业科技大学 | Panoramic type moving object surveillance method based on random update strategies |
-
2014
- 2014-01-15 CN CN201410017422.XA patent/CN103778629A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102314691A (en) * | 2011-06-30 | 2012-01-11 | 北京平安视讯科技有限公司 | Background model based on multiple information integration |
| CN102629383A (en) * | 2012-02-28 | 2012-08-08 | 湖南大学 | Motion object detection method based on random strategy |
| CN102665068A (en) * | 2012-04-26 | 2012-09-12 | 中南林业科技大学 | Panoramic type moving object surveillance method based on random update strategies |
Non-Patent Citations (1)
| Title |
|---|
| 胡雄鸽等: "使用随机策略进行运动目标检测方法研究", 《计算机工程与应用》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105184814A (en) * | 2015-07-27 | 2015-12-23 | 成都天奥信息科技有限公司 | Moving target detecting and tracking method based on multi-frame radar image |
| CN106023259A (en) * | 2016-05-26 | 2016-10-12 | 史方 | Method and device for detecting moving target frequency |
| CN106023259B (en) * | 2016-05-26 | 2018-12-07 | 史方 | A kind of moving target frequency detecting method and device |
| CN108491796A (en) * | 2018-03-22 | 2018-09-04 | 电子科技大学 | A kind of time domain period point target detecting method |
| CN108491796B (en) * | 2018-03-22 | 2021-10-22 | 电子科技大学 | Time domain periodic point target detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109919002B (en) | Yellow stop line identification method and device, computer equipment and storage medium | |
| CN107358258B (en) | SAR image target classification based on NSCT double CNN channels and selective attention mechanism | |
| CN103136766B (en) | A kind of object conspicuousness detection method based on color contrast and color distribution | |
| CN104463795B (en) | A kind of dot matrix DM image in 2 D code processing method and processing device | |
| CN102096821A (en) | Number plate identification method under strong interference environment on basis of complex network theory | |
| CN103268481A (en) | Method for extracting text in complex background image | |
| CN104899839A (en) | Ghost quick-inhibition method based on ViBe algorithm | |
| CN105868734A (en) | Power transmission line large-scale construction vehicle recognition method based on BOW image representation model | |
| Li et al. | A lane marking detection and tracking algorithm based on sub-regions | |
| CN104766344A (en) | Vehicle detecting method based on moving edge extractor | |
| CN106446921A (en) | High-voltage power transmission line barrier identification method and apparatus | |
| CN116311000A (en) | Firework detection method, device, equipment and storage medium | |
| CN110310263B (en) | SAR image residential area detection method based on significance analysis and background prior | |
| CN103778629A (en) | Background model real-time updating method for non-coherent radar image | |
| CN105354547A (en) | Pedestrian detection method in combination of texture and color features | |
| CN108010050B (en) | Foreground detection method based on adaptive background updating and selective background updating | |
| CN104537632A (en) | Infrared image histogram enhancing method based on edge extraction | |
| CN113393430A (en) | Thermal imaging image enhancement training method and device for fan blade defect detection | |
| Shi et al. | Image enhancement for degraded binary document images | |
| CN105205833A (en) | Moving object detection method and device based on space-time background model | |
| CN103093241B (en) | Based on the remote sensing image nonuniformity cloud layer method of discrimination of homogeneity process | |
| CN103034864A (en) | Video banner identification method based on color threshold and corner detection | |
| CN106339709A (en) | Real-time image extraction method | |
| CN102842025B (en) | The detection scene determination methods of video image and device | |
| CN110188601B (en) | Airport remote sensing image detection method based on learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140507 |
|
| RJ01 | Rejection of invention patent application after publication |