CN104168404B - A kind of thermal camera night vision antidote - Google Patents

A kind of thermal camera night vision antidote Download PDF

Info

Publication number
CN104168404B
CN104168404B CN201410359090.3A CN201410359090A CN104168404B CN 104168404 B CN104168404 B CN 104168404B CN 201410359090 A CN201410359090 A CN 201410359090A CN 104168404 B CN104168404 B CN 104168404B
Authority
CN
China
Prior art keywords
image
msub
video
night
mrow
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.)
Active
Application number
CN201410359090.3A
Other languages
Chinese (zh)
Other versions
CN104168404A (en
Inventor
王兴国
穆科明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NANJING JIEMAI VIDEO TECHNOLOGY Co Ltd
Original Assignee
NANJING JIEMAI VIDEO TECHNOLOGY Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by NANJING JIEMAI VIDEO TECHNOLOGY Co Ltd filed Critical NANJING JIEMAI VIDEO TECHNOLOGY Co Ltd
Priority to CN201410359090.3A priority Critical patent/CN104168404B/en
Publication of CN104168404A publication Critical patent/CN104168404A/en
Application granted granted Critical
Publication of CN104168404B publication Critical patent/CN104168404B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention discloses a kind of thermal camera night vision antidote, belong to HD video field of sensing technologies and technical field of video monitoring, especially one kind is related to infrared photography night vision adaptive corrective technology.This method obtains the standardized images of the Infrared jamming noise under certain image exposure parameter by measuring, and the image obtained during by it with night normal photographing be standardized after carry out calculus of differences, remove the interference noise of captured image, the scene image of the high-resolution high-contrast after being corrected after anti-standardization.The present invention can not change existing ball cover accessory, under the conditions of not increasing coating cost, by measuring and handling, and significantly improve night infrared illumination imaging effect so that image clearly after final process is exquisitely carved, high resolution.

Description

A kind of thermal camera night vision antidote
Technical field
The invention belongs to HD video field of sensing technologies and technical field of video monitoring, especially one kind is related to infrared photography Night vision correcting technology.
Background technology
Municipal public safety is the important leverage of urban sustainable development, and video monitoring is the important hand of municipal public safety Section, HD video video camera is the important development direction of field of video monitoring.In order to realize day and night round-the-clock monitor task, improve Night video acquisition quality, the day and night type video camera with infrared illumination obtains extensive popularization and application.Day and night round-the-clock monitoring It is made up of with video camera imaging sensor, optical lens, infrared illumination device LED and transparent ball cover, structure such as Fig. 1 institutes Show.
In order to realize fire prevention, explosion-proof, dust-proof, resistance to compression, attractive in appearance, easy processing, easy to install, the easy purpose, transparent ball cover such as use The general PC materials using semi-spherical shape.Thermal camera uses infrared LED active illumination when night natural lighting is not enough, The reflected light that infrared LED light is invested to object in scene, scene projects imaging sensor progress light by camera lens Electricity is converted into picture signal.
However, due to the characteristic of infrared illumination optics system so that nighttime image quality declines clearly with respect to daytime.
The transparent ball cover of actual use, general light transmittance is 90% or so.As shown in Fig. 2 night uses active infrared LED When illumination light shoots video image, infrared LED light is through PC ball covers to target illumination, and 90% or so light shines through PC ball covers It is mapped in target scene, lens imaging is reflected into body surface.Other about 10% light are reflected by ball cover, reflect or inhaled Receive, wherein 4% or so light energy reflects in the outer wall of ball cover, 2% or so light reflects along in PC materials, reflects and inhaled Receive, 4% or so light energy reflexes to camera lens in the inwall of ball cover, into lens imaging, form ambient interferences image.
Because the reflectivity of object scene in the case of night vision is not also high, thus cause object scene reflected light enter camera lens with The light intensity that interference light enters camera lens arrival sensor can be compared, and both are superimposed, and echo signal is flooded by ambient noise, so that Make the image blurring unclear, contrast of intake weak, actual scene image can not be seen clearly
With being continuously increased for video monitoring demand, the data in the urgent need to improving night monitoring video.The present invention is provided The problem of a kind of method for eliminating night infrared interference noise can be solved above well.
The content of the invention
The purpose of the present invention is:A kind of method for eliminating night infrared interference noise is provided, this method is obtained by measuring The standardized images of Infrared jamming noise under certain image exposure parameter, and the figure obtained during by it with night normal photographing As carrying out calculus of differences after being standardized, the interference noise of captured image is removed, after being corrected after anti-standardization High-resolution high-contrast scene image.
Specifically, the present invention is realized using following technical scheme:A kind of thermal camera night vision correction side Method, it is characterised in that set up the measuring system of infrared illumination interference, obtains the image of infrared illumination interference noise;At night During actual photographed, the illumination that infrared night vision is eliminated by interference compensation mechanism with correction structural system is disturbed.Object scene is anti- The image for penetrating light into camera lens is g (x, y), and the noise image that infrared illumination interference light enters camera lens arrival sensor formation is N (x, y), the image after both are superimposed is f (x, y), then has f (x, y)=g (x, y)+n (x, y), and noise pattern is obtained by measuring As n (x, y), by differential compensation computing, image g (x, y)=f (x, the y)-n (x, y) for removing interference, including following step are obtained Suddenly:
1) video camera is opened and is set to Night, and opens infrared LED lamp, real time shooting sequence of video images;
2) signal processing chip in video camera carries out " time orientation low pass spatial filtering " processing to sequence of video images, Test pattern component is filtered out, jamming pattern noise image is obtained;
3) using " image standardized processing algorithm " to noise image processing, and by the background noise graph of normalized processing As being recorded in video camera;
4) actual scene image sequence is absorbed, and is handled using " image standardized processing algorithm ";
5) the ambient noise image that the actual scene image of normalized processing is recorded in step 3 in video camera is carried out Differential compensation processing, removes jamming pattern image, obtains the actual scene image of standardization;
6) the actual scene image of the standardization obtained to step 5 is located using " image is counter to standardize Processing Algorithm " Reason, obtains removing the real image sequence of interference.
Above-mentioned technical proposal is further characterized by, when obtaining infrared illumination jamming pattern noise image, to step 2 The sequence of video images of intake carries out " time orientation low pass spatial filtering ", and its form is shown below:
Wherein fl(x, y) is wave filter output, f (n1, n2) it is view data to be filtered, pixel value is taken by time orientation Square formation is constituted, filter function H is L × L square formations, generally using 3x3 LPF operator, so f (n1, n2) with three two field pictures For a processing unit, each frame takes three pixels of same a line, the progress of 3x3 squares is made up of the pixel of adjacent three rows same position Low-pass filtering treatment, to realize time orientation LPF.
Above-mentioned technical proposal is further characterized by, and " image standardized processing algorithm " comprises the following steps:
1) image histogram is determined, it is given below to seek kth ash if the tonal gradation scope of filtered image is [0, M-1] Spend the function of the number of pixels of level
h(rk)=nk
Wherein k=0,1 ..., M-1 are kth level gray scale, nkRepresent that gray level is r in imagekNumber of pixels;
2) determine that minimum pixel value is m in histograml, maximum is mh, the following tercile of calculating histogram (10%, m1), (20%, m2) ..., (90%, m9), wherein, miFor the corresponding pixel value of hundredths;
3) pixel value of interesting image scope is set as slAnd sh, by mlAnd mhBetween mi be mapped to s by space ratiolWith shBetween and round, mlAnd mhIt is mapped as slAnd sh
4) linear transformation is carried out to every section between each tercile, completes image standardized processing.
Above-mentioned technical proposal is further characterized by, and " image standardized processing algorithm " comprises the following steps:
1) image histogram is determined;
2) minimum pixel value is set in histogram as sl, maximum is sh, calculate histogram following tercile (10%, p1)、 (20%, p2) ..., (90%, p9), wherein, piFor the corresponding pixel value of hundredths;
3) set image actual pixels scope and be set to glAnd gh, by slAnd shBetween piG is mapped to by space ratiolAnd gh Between and round, be slAnd shIt is mapped as glAnd gh
4) linear transformation is carried out to every section between each tercile, completes the anti-standardization of image.
Above-mentioned technical proposal is further characterized by, and the acquisition methods of intake test video image sequence include providing one The test rotating disk with special pattern is planted, its video image is obtained to be not less than when 50 revolution per second speed are rotated in rotating disk.
Beneficial effects of the present invention are as follows:To overcome interference of the infrared illumination to image, this hair in infrared night video camera Bright actual measurement and processing by Infrared jamming, it is removed from actual video shooting, it is hereby achieved that high-resolution Infrared night vision imaging.Compared with existing infrared imaging system, existing ball cover accessory can not changed, do not increase plated film into Under the conditions of this, by measuring and handling, night infrared illumination imaging effect is significantly improved so that the image clearly after final process Exquisitely carved, high resolution.
Brief description of the drawings
Fig. 1 is infrared day and night type video camera photoelectricity structure chart.
Fig. 2 is camera optics imaging and interference figure picture formation figure.
Fig. 3 is image histogram and brightness tercile chart.
Fig. 4 is image standardization conversion schematic diagram.
Fig. 5 is the anti-standardized transformation schematic diagram of image.
Fig. 6 is Infrared jamming image capture flow figure.
Fig. 7 is a kind of test pattern rotating disk.
Fig. 8 bends overlay area is the video camera imaging visual field.
Fig. 9 is real image intake process flowchart.
Embodiment
The present invention is described in further detail with reference to Figure of description.
Vision signal time orientation spatial low pass wave method:
When obtaining infrared illumination interference noise image, spatial domain low pass filtered is carried out on time orientation to the video of intake Ripple, its form is shown below:
Wherein fl(x, y) is wave filter output, f (n1, n2) it is view data to be filtered, pixel value is taken by time orientation Square formation is constituted, filter function H is L × L square formations, generally using 3x3 LPF operator, so f (n1, n2) with three two field pictures For a processing unit, each frame takes three pixels of same a line, the progress of 3x3 squares is made up of the pixel of adjacent three rows same position Low-pass filtering treatment, to realize time orientation LPF.
Image histogram:
If the tonal gradation scope of filtered image is [0, M-1], formula (2) provides the number of pixels of seeking kth gray level Function
h(rk)=nk (2)
Wherein k=0,1 ..., M-1 are kth level gray scale, nkRepresent that gray level is r in imagekNumber of pixels.
Image standardized processing algorithm:
1. image histogram is determined, as shown in Figure 3;
2. determine that minimum pixel value is m in histograml, maximum is mh, the following tercile of calculating histogram (10%, m1), (20%, m2) ..., (90%, m9), wherein, miFor the corresponding pixel value of hundredths;
3. the pixel value of interesting image scope is set as slAnd sh, by mlAnd mhBetween mi be mapped to s by space ratiolWith shBetween and round, mlAnd mhIt is mapped as slAnd sh
4. linear transformation is carried out to every section between each tercile, as shown in figure 4, completing image standardized processing.
Image is counter to standardize Processing Algorithm:
1. determine image histogram;
2. minimum pixel value is set in histogram as sl, maximum is sh, calculate histogram following tercile (10%, p1)、 (20%, p2) ..., (90%, p9), wherein, piFor the corresponding pixel value of hundredths;
3. setting image actual pixels scope is set to glAnd gh, by slAnd shBetween piG is mapped to by space ratiolAnd gh Between and round, be slAnd shIt is mapped as glAnd gh
4. between each tercile to every section of progress linear transformation, as shown in figure 5, completing image instead at standardization Reason.
The method that the present invention eliminates night infrared interference noise has two key steps, including:
Step one, Infrared jamming noise video is obtained;This step comprises the following steps (see Fig. 6)
1. the test image rotating disk of pattern shown in Fig. 7, video camera are placed with optical axis direction in camera coverage The region for absorbing image is as shown in Figure 8;
2. video camera is opened and is set to Night, and opens infrared LED lamp;
3. it is adjusted to as exposure parameter to appropriate value;
4. rotate rotating disk to be not less than 50 revolution per second speed;
5. video camera real time shooting sequence of video images;
6. the signal processing chip in video camera carries out " time orientation low pass spatial filtering " processing to sequence of video images, Test pattern component is filtered out, jamming pattern noise image is obtained;
7. using " image standardized processing algorithm " to noise image processing;
8. the ambient noise image of normalized processing is recorded in video camera.
Step 2, following operating procedure is performed when actual scene is shot (see Fig. 9):
1. video camera is opened and is set to Night, and opens infrared LED lamp;
2. it is adjusted to as exposure parameter to appropriate value;
3. the actual scene image sequence of intake;
4. using " image standardized processing algorithm " to image procossing;
5. the actual scene image of pair normalized processing and the ambient noise image being recorded in video camera carry out difference Compensation deals, remove jamming pattern image, obtain the actual scene image of standardization;
6. pair step 5 gained image is handled using " image is counter to standardize Processing Algorithm ", obtain removing the reality of interference Border image sequence.
Although the present invention is disclosed as above with preferred embodiment, embodiment is not for limiting the present invention's.Not In the spirit and scope for departing from the present invention, any equivalence changes done or retouching also belong to the protection domain of the present invention.Cause The content that this protection scope of the present invention should be defined using claims hereof is standard.

Claims (3)

1. a kind of thermal camera night vision antidote, it is characterised in that methods described be by infrared illumination is disturbed into Row measurement, obtains the noise image of infrared illumination interference, and in night actual photographed video, is eliminated by interference compensation red The illumination interference of outer night vision;The video image that object scene reflected light enters camera lens is g (x, y), and infrared illumination interference light is entered The noise image for entering camera lens arrival sensor formation is n (x, y), and the image after both are superimposed is f (x, y), then has f (x, y)=g (x, y)+n (x, y), noise image is obtained for n (x, y) by measuring, and by differential compensation computing, obtains the video for removing interference Image g (x, y)=f (x, y)-n (x, y), comprises the following steps:
1) video camera is opened and is set to Night, and opens infrared LED lamp, real time shooting test video image sequence;
2) signal processing chip in video camera carries out " time orientation low pass spatial filtering " processing to sequence of video images, filters out Test pattern component, obtains jamming pattern noise image;
3) using " image standardized processing algorithm " to noise image processing, and the ambient noise image of normalized processing is remembered Record in video camera, described image standardized processing algorithm comprises the following steps:
Image histogram 3-1) is determined, it is given below to seek kth gray scale if the tonal gradation scope of filtered image is [0, M-1] The function of the number of pixels of level:
h(rk)=nk
Wherein k=0,1 ..., M-1 are kth level gray scale, nkRepresent that gray level is r in imagekNumber of pixels;
3-2) determine that minimum pixel value is m in histograml, maximum is mh, calculate histogram following tercile (10%, m1)、 (20%, m2) ..., (90%, m9), wherein, miFor the corresponding pixel value of hundredths;
The pixel value of interesting image scope 3-3) is set as slAnd sh, by mlAnd mhBetween miS is mapped to by space ratiolAnd sh Between and round, mlAnd mhIt is mapped as slAnd sh
Linear transformation 3-4) is carried out to every section between each tercile, image standardized processing is completed;
4) actual scene image sequence is absorbed, and is handled using " image standardized processing algorithm ";
5) to the actual scene image of normalized processing and step 3) the ambient noise image that is recorded in video camera carry out it is poor Divide compensation deals, remove jamming pattern image, obtain the actual scene image of standardization;
6) to step 5) the actual scene image of standardization that obtains handled using " image is counter to standardize Processing Algorithm ", obtained To the real image sequence for removing interference, the anti-Processing Algorithm that standardizes of described image comprises the following steps:
6-1) determine image histogram;
Minimum pixel value 6-2) is set in histogram as sl, maximum is sh, calculate histogram following tercile (10%, p1)、 (20%, p2) ..., (90%, p9), wherein, piFor the corresponding pixel value of hundredths;
6-3) set image actual pixels scope and be set to glAnd gh, by slAnd shBetween piG is mapped to by space ratiolAnd ghIt Between and round, be slAnd shIt is mapped as glAnd gh
Linear transformation 6-4) is carried out to every section between each tercile, the anti-standardization of image is completed.
2. a kind of thermal camera night vision antidote according to claim 1, it is characterised in that obtaining infrared illumination During jamming pattern noise image, to step 2) intake sequence of video images carry out " time orientation low pass spatial filtering ", its shape Formula is shown below:
<mrow> <msub> <mi>f</mi> <mi>l</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mo>&amp;Sigma;</mo> <msub> <mi>n</mi> <mn>1</mn> </msub> </munder> <munder> <mo>&amp;Sigma;</mo> <msub> <mi>n</mi> <mn>2</mn> </msub> </munder> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>n</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>n</mi> <mn>2</mn> </msub> <mo>)</mo> </mrow> <mi>H</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>-</mo> <msub> <mi>n</mi> <mn>1</mn> </msub> <mo>+</mo> <mn>1</mn> <mo>,</mo> <mi>y</mi> <mo>-</mo> <msub> <mi>n</mi> <mn>2</mn> </msub> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow>
Wherein fl(x, y) is wave filter output, f (n1, n2) it is view data to be filtered, take pixel value to constitute by time orientation Square formation, filter function H is L × L square formations, generally using 3x3 LPF operator, so f (n1, n2) using three two field pictures as one Processing unit, each frame takes three pixels of same a line, 3x3 squares progress low pass is made up of the pixel of adjacent three rows same position Filtering process, to realize time orientation LPF.
3. a kind of thermal camera night vision antidote according to claim 1, it is characterised in that intake test video figure As the acquisition methods of sequence include providing a kind of test rotating disk with special pattern, in rotating disk to be not less than 50 revolution per second speed Its video image is obtained during rotation.
CN201410359090.3A 2014-07-25 2014-07-25 A kind of thermal camera night vision antidote Active CN104168404B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410359090.3A CN104168404B (en) 2014-07-25 2014-07-25 A kind of thermal camera night vision antidote

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410359090.3A CN104168404B (en) 2014-07-25 2014-07-25 A kind of thermal camera night vision antidote

Publications (2)

Publication Number Publication Date
CN104168404A CN104168404A (en) 2014-11-26
CN104168404B true CN104168404B (en) 2017-09-12

Family

ID=51912016

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410359090.3A Active CN104168404B (en) 2014-07-25 2014-07-25 A kind of thermal camera night vision antidote

Country Status (1)

Country Link
CN (1) CN104168404B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108305378A (en) * 2017-01-11 2018-07-20 广州广电运通金融电子股份有限公司 A kind of image processing method, device and system

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN2842939Y (en) * 2005-10-28 2006-11-29 沈阳理工大学 Anti-fog anti-interference pick-up device
US7928395B2 (en) * 2003-10-02 2011-04-19 Opgal Ltd. Infra-red (IR) sensor with controllable sensitivity
CN102051871A (en) * 2010-10-21 2011-05-11 东华大学 Embedded watercourse floating pollutant vision detecting device
CN102262775A (en) * 2010-05-26 2011-11-30 深圳迈瑞生物医疗电子股份有限公司 Image denoising method and device and image processing equipment
CN102665034A (en) * 2012-03-16 2012-09-12 江苏华枫物联网科技有限公司 Night effect removal method for camera-collected video
CN103500341A (en) * 2013-09-16 2014-01-08 安徽工程大学 Recognition device used for road signboard
CN103845077A (en) * 2012-12-05 2014-06-11 深圳迈瑞生物医疗电子股份有限公司 Gain optimization method for ultrasound images and ultrasound imaging gain automatic optimization device

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1484042A (en) * 2003-07-03 2004-03-24 上海交通大学 Energy projection detection method for slow action small and weak object in infrared image sequence

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7928395B2 (en) * 2003-10-02 2011-04-19 Opgal Ltd. Infra-red (IR) sensor with controllable sensitivity
CN2842939Y (en) * 2005-10-28 2006-11-29 沈阳理工大学 Anti-fog anti-interference pick-up device
CN102262775A (en) * 2010-05-26 2011-11-30 深圳迈瑞生物医疗电子股份有限公司 Image denoising method and device and image processing equipment
CN102051871A (en) * 2010-10-21 2011-05-11 东华大学 Embedded watercourse floating pollutant vision detecting device
CN102665034A (en) * 2012-03-16 2012-09-12 江苏华枫物联网科技有限公司 Night effect removal method for camera-collected video
CN103845077A (en) * 2012-12-05 2014-06-11 深圳迈瑞生物医疗电子股份有限公司 Gain optimization method for ultrasound images and ultrasound imaging gain automatic optimization device
CN103500341A (en) * 2013-09-16 2014-01-08 安徽工程大学 Recognition device used for road signboard

Also Published As

Publication number Publication date
CN104168404A (en) 2014-11-26

Similar Documents

Publication Publication Date Title
Stumpfel et al. Direct HDR capture of the sun and sky
CN107274336B (en) A kind of Panorama Mosaic method for vehicle environment
CN106572340B (en) Camera system, mobile terminal and image processing method
CN102177719B (en) Apparatus for detecting direction of image pickup device and moving body comprising same
US20170257549A1 (en) Optical filter opacity control for reducing temporal aliasing in motion picture capture
JP5545016B2 (en) Imaging device
TW201230773A (en) Image capture using three-dimensional reconstruction
CN101422035A (en) Image high-resolution upgrading device, image high-resolution upgrading method, image high-resolution upgrading program and image high-resolution upgrading system
Raskar et al. Computational photography
CN110519489A (en) Image-pickup method and device
CN107079087A (en) Camera device and object recognition methods
CN108712608A (en) Terminal device image pickup method and device
Seckiner et al. Forensic image analysis–CCTV distortion and artefacts
CN107392192A (en) A kind of personal identification method, device and multi-cam module
CN108776980A (en) A kind of scaling method towards lenticule light-field camera
Sen et al. Practical high dynamic range imaging of everyday scenes: Photographing the world as we see it with our own eyes
WO2021077944A1 (en) Method for eliminating periodic noise in light field reconstructed image based on frequency domain
CN113554575B (en) High-reflection object surface highlight removing method based on polarization principle
CN106023193A (en) Array camera observation method for detecting structure surface in turbid media
CN110490811A (en) Image noise reduction apparatus and image denoising method
CN102457669A (en) Image processing method
CN104168404B (en) A kind of thermal camera night vision antidote
CN110827375B (en) Infrared image true color coloring method and system based on low-light-level image
CN108600631A (en) Image processing method, device, computer readable storage medium and electronic equipment
WO2016147782A1 (en) Film thickness measurement device and film thickness measurement method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant