CN105894500B - A kind of visual range detection method based on image procossing - Google Patents
A kind of visual range detection method based on image procossing Download PDFInfo
- Publication number
- CN105894500B CN105894500B CN201610185869.7A CN201610185869A CN105894500B CN 105894500 B CN105894500 B CN 105894500B CN 201610185869 A CN201610185869 A CN 201610185869A CN 105894500 B CN105894500 B CN 105894500B
- Authority
- CN
- China
- Prior art keywords
- visibility
- visual range
- image
- thermal
- visual
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Radiation Pyrometers (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The visual range detection method based on image procossing that the present invention relates to a kind of, this method comprises the following steps：(1) it acquires normal image under current scene and obtains edge pixel point sum；(2) thermal-induced imagery edge pixel point sum and corresponding thermal-induced imagery correction factor under corresponding scene are obtained；(3) visibility model is established；(4) the visual angle value under current scene is calculated according to visibility model, and the visibility consult table for comparing storage obtains corresponding visual range, wherein including the mapping relations under different scenes between visibility standard value and visual range in the visibility consult table stored.Compared with prior art, the present invention has many advantages, such as that method is simple, image quality requirements are low, accuracy of detection is high.
Description
Technical field
The present invention relates to a kind of visual range detection methods, are examined more particularly, to a kind of visual range based on image procossing
Survey method.
Background technology
Turnpike driving visual range is an important factor for influencing traffic safety, and traditional visibility Testing index is easy
By influences such as illumination condition, environment temperature, sleet, it is difficult to the practical visual range of driver is accurately reflected, and visibility
Instrument is expensive, erection mode is complicated, is used it is difficult to lay on a large scale.
The Video Detection Algorithm research of existing low visibility is mainly pixel contrast algorithm, using as defined in CIE 0.05
Human eye visible contrast threshold value, including：Computational methods, the phase of computational methods, visibility meter correction based on data correction
The computational methods of machine self-calibration；In addition to this, the method also detected into line visibility based on road surface brightness, these methods are all
It is to be analyzed for normal image itself, therefore it is higher to image quality requirements, when the quality of acquired image is relatively low
Just without research significance, while the appropriate processing of image and certain insufficient (such as hardware is delayed) of associated hardware circuitry are also such
The bottleneck of method.
Invention content
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind being based on image procossing
Highway visual range detection method.
The purpose of the present invention can be achieved through the following technical solutions：
A kind of visual range detection method based on image procossing, this method comprises the following steps：
(1) it acquires normal image under current scene and obtains normal image edge pixel point sum N；
(2) thermal-induced imagery edge pixel point sum n ' and corresponding thermal-induced imagery amendment system under corresponding scene are obtained
Number β；
(3) visibility model is establishedWherein v is visual angle value；
(4) the visual angle value under current scene is calculated according to visibility model, and the visibility consult table for comparing storage obtains
Corresponding visual range is taken, wherein including visibility standard value and visual range under different scenes in the visibility consult table stored
Between mapping relations.
The visibility consult table of the storage is obtained by following manner：The common figure under different scenes is obtained respectively
Picture is respectively processed the normal image under different scenes, and the mapping established between visual range and visibility standard value is closed
System, and then establish the visibility consult table for including different scenes.
The mapping relations established between visual range and visibility standard value are specially：
(1a) determines visual range separation x0、x1、x2……xn, wherein x0=0m, x0< x1< x2……xn；
(1b) extracts visual range range (x in normal imagej-1, xj) in scenery and detached from the normal image
Form secondary image Pj, j=1,2 ... n；
(1c) is to secondary image PjIt carries out edge detection and obtains the edge pixel points Q of the secondary imagej；
(1d) is to visual range separation xiVisibility standard value is sought according to the following formula：
Wherein, i=1,2 ... n, as i=0, visual range separation x0Visibility standard value v0=0.
Normal image and thermal-induced imagery carry out the edge that Image Edge-Detection obtains each image by Canny operators respectively
Pixel sum.
Correction factor β=α τ in step (2), wherein α are temperature correction coefficient, and τ is camera correction factor.
Temperature correction coefficient α is obtained by following formula：
Wherein, C1For first radiation constant, C2For second radiation constant, λ is thermal infrared imager operation wavelength, TiFor shooting
Temperature when thermal-induced imagery, T are reference temperature, TiIt is Kelvin with T unit.
Camera correction factor τ is obtained by following methods：
(2a) obtains m normal image and corresponding thermal-induced imagery under same position and same view angle；
The edge detection pixel average value that (2b) seeks m normal image respectively is R, correspondingly, m thermal-induced imagery
Edge detection pixel average value be S；
(2c) calculates camera correction factor
The step (4) is specially：
Visual angle value is calculated, the visibility standard value whether having equal to the visual angle value in visibility consult table is compared,
If so, then obtaining the corresponding visual range of visibility standard value, it is corresponding otherwise to seek visual angle value by linear interpolation
Visual range.
Compared with prior art, the invention has the advantages that：
(1) by image processing techniques come the edge pixel point sum of detection image, then to the pixel of thermal-induced imagery
Camera amendment and temperature adjustmemt are carried out, is come finally by the pixel number of normal image and the ratio of thermal-induced imagery correction result
It is counter to push away practical visual range, i.e., reflect the decrement of visual range by the decrement of scenery in the visual field, passes through infrared chart
As simplifying image processing algorithm come mode compare with normal image, requirement of the reduction to picture quality；
(2) visibility consult table is the visibility consult table for making and having stored, therefore when actually detected, only needs to call
The visibility consult table selects corresponding scene and can carry out the access of visual angle value, easy to use；
(3) visibility detection model is simple, and when application need to only acquire normal image and calculate edge pixel point sum N, and
Thermal-induced imagery edge pixel point sum n ' and correction factor can be previously stored under corresponding scene, under Same Scene,
Whether which kind of weather, thermal-induced imagery edge pixel point sum n ' and correction factor β are remained unchanged, this mode side of processing
Just, it is easy to use；
(4) temperature correction coefficient reduces temperature to infrared image influence in the invention, while camera correction factor reduces
Thermal infrared imager image quality difference, effectively increases the computational accuracy of visibility model, to improve accuracy of detection.
Description of the drawings
Fig. 1 is that the present invention is based on the flow charts of the visual range detection method of image procossing；
Fig. 2 is the linear fit curve comparison figure of the pixel number before and after thermal-induced imagery temperature adjustmemt.
Specific implementation mode
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.
Embodiment
A kind of simple model algorithm that the purpose of the present invention proposes, only detection are less than the low visibility range of 300m, and
And propose the visibility concept for being specially adapted for highway, which is primarily upon driver's institute's view as reflection is true
Human viewable's distance.The final detection method that a kind of customized economy is detected for highway visual range, and realize real-time
Detection early warning.
The present invention basic principle thinking be：
Low visibility is exactly visual range deficiency to directly affecting for driver, and visual range is smaller, and human eye can be seen
Farthest scenery distance with regard to smaller, the scenery quantity in normal image is also fewer.So with the reduction of visual range, continuously
The amount of scenery regular will be reduced.Therefore visual range can be pushed away come counter by quantifying the amount of scenery reduction in normal image
Decrement, obtain practical visual range；Quantify the scenery amount of the reduction in normal image because of visual range and reduction, it is first
Thermal-induced imagery is first introduced as reference.Infra-red radiation is the main factor for influencing thermal-induced imagery imaging, therefore infrared heat
Image does not change with the variation of visual range, is not influenced by the low visual range condition such as haze, can be used as and weigh commonly
The reference of image scene decrement.Through overtesting as can be seen that the scenery quantity in the normal image in greasy weather obviously subtracts compared with fine day
It is few, and thermal-induced imagery is substantially unchanged；Then, it needs to introduce the image parameter that can reflect scenery quantity, using scenery side
Edge pixel number is as this parameter.Each object corresponds to the edge pixel of oneself in image, and scenery gets at most total side
Edge pixel is more；Specific how to quantify in normal image the scenery amount of reduction due to visual range is small, the present invention is using common
A small amount of index is seen as scenery with the ratio of the edge detection pixel number of thermal-induced imagery.Ratio is 1, illustrates normal image
Scenery quantity is consistent with thermal-induced imagery scenery quantity, and visual range is very high, and ratio is reduced less than 1 scenery quantity, can
The apparent distance reduces；After taking pixel ratio index, the mapping relations for establishing practical visual range and the ratio are needed, according to this
Mapping relations can push away practical visual range by the edge detection pixel ratio actually calculated come counter.The present invention is by building
Visibility consult table is found to reflect the mapping relations of visual range and edge detection pixel ratio.
It is therefore proposed that a kind of highway visual range detection method based on image procossing, this method flow chart such as Fig. 1
It is shown, include the following steps：
Execute step 1：It acquires normal image under current scene and obtains normal image edge pixel point sum N；
Execute step 2：Obtain thermal-induced imagery edge pixel point sum n ' and corresponding infrared chart under corresponding scene
As correction factor β；
Execute step 3：Establish visibility modelWherein v is visual angle value；
Execute step 4：The visual angle value under current scene is calculated according to visibility model, and the visibility for comparing storage is looked into
It reads table and obtains corresponding visual range, wherein comprising visibility standard value under different scenes and can in the visibility consult table stored
Mapping relations between the apparent distance.
Specifically, normal image and thermal-induced imagery carry out Image Edge-Detection by Canny operators respectively obtains each figure
The edge pixel point sum of picture.The visibility consult table of storage is obtained by following manner：It obtains respectively general under different scenes
Logical image, handles normal image, establishes the mapping relations between visual range and visibility standard value, and then establish packet
Visibility consult table containing different scenes.The mapping relations wherein established between visual range and visibility standard value are specially：
(1a) determines visual range separation x0、x1、x2……xn, wherein x0=0m, x0< x1< x2……xn；
(1b) extracts visual range range (x in normal imagej-1, xj) in scenery and detached from the normal image
Form secondary image Pj, j=1,2 ... n；
(1c) is to secondary image PjIt carries out edge detection and obtains the edge pixel points Q of the secondary imagej；
(1d) is to visual range separation xiVisibility standard value is sought according to the following formula：
Wherein, i=1,2 ... n, as i=0, visual range separation x0Visibility standard value v0=0.
Monochromatic radiation emittance when black matrix is in temperature T in af at wavelength lambda is provided by planck formula, can be derived from certain
Wave-length coverage in energy integral be formula：
In formula, C1For first radiation constant, C2For second radiation constant.Differential declines wave plate is installed additional with reference to pertinent literature
Non-refrigeration focal surface infrared thermal imagery instrument temperature and energy relationship curve, by curve it is found that in the low temperature range within 300k
Non-refrigeration focal surface thermal infrared imager temperature is parabola, as quadratic relationship with energy relationship curve approximation.Set temperature becomes
Change range and be no more than 10 DEG C, and temperature, all within 300k, amplitude of variation is smaller, it can be assumed that the picture that infrared image processing obtains
Vegetarian refreshments number is quadratic relationship with temperature.Therefore, it can set on the basis of 25 DEG C of infrared image, correction factor 1.00,
Temperature correction coefficient so within the scope of 18 DEG C~28 DEG C of experimental temperature can be expressed such as formula：
Wherein, C1=2 π hc2=3.74 × 10-16Wm2, second radiation constant 0.01438769mK, λ are wavelength, here
It is 8~14 μm to take thermal infrared imager operation wavelength, thermal imager operation wavelength, can take 10 μm of intermediate value, TiTo shoot infrared chart
As when temperature, T is reference temperature, i.e. TiIt is Kelvin with T unit, is selected as 25 DEG C herein with reference to temperature, corresponding Kelvin
Temperature 298.16K.
Camera correction factor τ is obtained by following methods：
(2a) obtains m normal image and corresponding thermal-induced imagery under same position and same view angle；
The edge detection pixel average value that (2b) seeks m normal image respectively is R, correspondingly, m thermal-induced imagery
Edge detection pixel average value be S；
(2c) calculates camera correction factor
Step 4 is specially：Calculate visual angle value, compare in visibility consult table whether have equal to the visual angle value can
Otherwise diopter standard value is sought visually if so, then obtaining the corresponding visual range of visibility standard value by linear interpolation
The corresponding visual range of angle value.
In order to which the correctness and reliability of verifying this method have carried out some experiments.Under different scenes, respectively in phase
With using camera and thermal infrared imager to obtain normal image and thermal-induced imagery under position and same view angle, specifically, by Sony
Digital camera DSC-W630 types card camera and the vehicle-mounted thermal imaging systems of Uncooled FPA detector EX-25 are set up in same position
It sets, keeps same view angle, be chosen under three kinds of greasy weather, rainy day and night low-light (level) scenes and carry out normal image and thermal-induced imagery
Acquisition, instrument is assumed in school of Jiading District in Shanghai City, carries out Image Acquisition under the conditions of above-mentioned three kinds respectively, and record
Temperature when each sample collection obtains totally eight groups of samples.In addition obtained normal image and thermal-induced imagery are dropped
It makes an uproar processing, thermal-induced imagery takes wavelet de-noising and medium filtering, normal image to take medium filtering.
The normal image under Same Scene is handled respectively according to the above method, establishes visual range and visual scale
Mapping relations between quasi- value.When being tested, determine visual range separation 0m, 50m, 100m, 150m, 200m, 250m,
Normal image under three kinds of scenes is separated into 6 secondary images, is denoted as P by 300m respectivelyj, j=1,2 ... 6, secondary image P1
It is the secondary image for including scenery in 0~50m, secondary image P2It is the secondary image for including scenery in 50m~100m, it is right
Visibility standard value is calculated separately in each visual range separation.One by one by above-mentioned visual range separation and visibility standard value
It is corresponding, the visibility consult table under different scenes is made, visibility consult table shown in table 1 is obtained, empty portions indicate in table
Without scenery within the scope of the corresponding visual range.
1 visibility consult table of table
Fig. 2 is to shoot infrared image at different temperatures, the linear fit curve comparison of pixel number before and after temperature adjustmemt
Scheme, dotted line is the linear fit curve for correcting preceding pixel points in figure, is represented by：Y=-442.22x+21721, solid line are to repair
The linear fit curve of positive after-image vegetarian refreshments number, is represented by：Y=-107.06x+13385.
The slope absolute value of fitting a straight line is decreased obviously after amendment as seen from Figure 2, reduce temperature to infrared image at
The influence for managing result, ensure that infrared image edge pixel number is almost the same under same visual angle and position.
Then the acquisition of camera correction factor τ is carried out：It is chosen under the scene of high-visibility and is tested, ensure common figure
Scenery in picture and infrared image is completely the same, obtains 10 groups of contrast images, detects the side of normal image and infrared image respectively
Edge pixel number, finally calculate normal image pixel number average value and the ratio of infrared image pixel number average value are
Camera correction factor, analysis result is as shown in table 2, and camera correction factor τ is the camera amendment considered after temperature adjustmemt in the table
Coefficient, environment temperature when obtaining 10 groups of images due to the experiment are 23.6 DEG C, the temperature equally on the basis of 25 DEG C, at this time temperature
Correction factor is 0.9630, thus by table 28036 divided by 14536 be equal to 1.9287 be temperature adjustmemt before camera amendment system
Number, by its multiplied by the camera correction factor just obtained with 0.9630 in table be 1.86.
2 thermal-induced imagery camera correction analysis result of table
Then a series of processing are carried out to eight groups of samples of acquisition to obtain visual range and carry out pair with human eye observation's distance
Than obtaining experimental result shown in table 3.Due to the thermal-induced imagery that under same scene, theoretically thermal infrared imager measures
It answers identical, but due to apparatus measures error problem, can have some deviations, therefore under Same Scene, several groups of samples carry out figure
After correcting, the revised edge pixel point sum of several groups of samples is taken average as visibility model calculation formula under the scene
In denominator, can improve in this way measure precision.
As can be known from Table 3, the practical human eye observation's visual range of rainy day environment is all higher than 300m, according to visibility model prediction
As a result and it is all higher than 300m, prediction result is accurate, and visibility model is reliable；Foggy environment, sample F2 visibility model predictions
As a result it is consistent with human eye observation, sample F1 visibility model prediction results have the gap of 17m, relative error 5.7%；Low-light (level) ring
Border, ambient light illumination when sample L2 is more some higher than sample L1, visual range also should bigger, practical human eye observed result
L2 is maximum, and visibility model prediction result is consistent with human eye observation's result, average relative error 9.7%.
The case where visibility model prediction ranging from 0~300m, prediction result is more than 300m and is consistent with actual conditions
Error is not calculated, therefore considers that visual range is less than 300m situations, model prediction result and human eye observation's result under various environment
Average relative error is 8.7%.
3 experimental result of table
The visual range that above-mentioned visual range detection method is applied to highway detects, it is necessary first to before doing some
Sequence works：First：Visibility consult table is established, specially freeway surveillance and control camera is utilized to acquire normal image, and establish
Visibility consult table under different scenes, the visibility consult table can be preserved and be used for follow-up；Second：Corresponding different field
It under scape, acquires thermal-induced imagery by manually carrying thermal imaging system and calculates thermal-induced imagery edge pixel point sum n, while really
Determine temperature correction coefficient and camera correction factor, these parameters can be preserved, can directly be used when using next time.Exist later
When actually detected, using freeway surveillance and control camera acquire normal image, the monitoring camera shooting angle need with before
Thermal-induced imagery shooting angle be consistent, the normal image under corresponding scene is taken by monitoring camera and is calculated
Edge pixel point sum N, is then calculated visual angle value v by visibility model, and finally being obtained by visibility consult table can
The apparent distance.One thermal infrared imager can certainly be installed in freeway surveillance and control camera, and keep identical installation site
Acquisition thermal-induced imagery in real time is carried out with identical shooting visual angle, such cost can be relatively high.
Claims (3)
1. a kind of visual range detection method based on image procossing, which is characterized in that this method comprises the following steps：
(1) it acquires normal image under current scene and obtains normal image edge pixel point sum N；
(2) thermal-induced imagery edge pixel point sum n ' and corresponding thermal-induced imagery correction factor β under corresponding scene is obtained,
Wherein, correction factor β=α τ, α are temperature correction coefficient, and τ is camera correction factor；
(3) visibility model is establishedWherein v is visual angle value；
(4) the visual angle value under current scene is calculated according to visibility model, and the visibility consult table for comparing storage obtains phase
The visual range answered, wherein comprising under different scenes between visibility standard value and visual range in the visibility consult table stored
Mapping relations；
Temperature correction coefficient α is obtained by following formula：
Wherein, C1For first radiation constant, C2For second radiation constant, λ is thermal infrared imager operation wavelength, TiIt is infrared to shoot
Temperature when thermal image, T are reference temperature, TiIt is Kelvin with T unit；
Camera correction factor τ is obtained by following methods：
(2a) obtains m normal image and corresponding thermal-induced imagery under same position and same view angle；
The edge detection pixel average value that (2b) seeks m normal image respectively is R, correspondingly, the side of m thermal-induced imagery
Edge detection pixel point average value is S；
(2c) calculates camera correction factor
The visibility consult table of the storage is obtained by following manner：The normal image under different scenes is obtained respectively, it is right
Normal image under different scenes is respectively processed, and establishes the mapping relations between visual range and visibility standard value, into
And establish the visibility consult table for including different scenes；
The mapping relations established between visual range and visibility standard value are specially：
(1a) determines visual range separation x0、x1、x2、……、xn, wherein x0=0 meter, x0< x1< x2< ... < xn；
(1b) extracts visual range range (x in normal imagej-1, xj) in scenery and from the normal image separation formed two
Secondary image Pj, j=1,2 ... ..., n；
(1c) is to secondary image PjIt carries out edge detection and obtains the edge pixel points Q of the secondary imagej；
(1d) is to visual range separation xiVisibility standard value is sought according to the following formula：
Wherein, i=1,2 ... ..., n, as i=0, visual range separation x0Visibility standard value v0=0.
2. a kind of visual range detection method based on image procossing according to claim 1, which is characterized in that common figure
Picture and thermal-induced imagery carry out the edge pixel point sum that Image Edge-Detection obtains each image by Canny operators respectively.
3. a kind of visual range detection method based on image procossing according to claim 1, which is characterized in that described
Step (4) is specially：
Visual angle value is calculated, the visibility standard value whether having equal to the visual angle value in visibility consult table is compared, if so,
The corresponding visual range of visibility standard value is then obtained, otherwise seeking that visual angle value is corresponding by linear interpolation can sighting distance
From.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610185869.7A CN105894500B (en) | 2016-03-29 | 2016-03-29 | A kind of visual range detection method based on image procossing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610185869.7A CN105894500B (en) | 2016-03-29 | 2016-03-29 | A kind of visual range detection method based on image procossing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105894500A CN105894500A (en) | 2016-08-24 |
| CN105894500B true CN105894500B (en) | 2018-10-26 |
Family
ID=57013866
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610185869.7A Active CN105894500B (en) | 2016-03-29 | 2016-03-29 | A kind of visual range detection method based on image procossing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105894500B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112730345A (en) * | 2019-10-28 | 2021-04-30 | 浙江宇视科技有限公司 | Visibility detection method, device, equipment and medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE29811086U1 (en) * | 1997-08-22 | 1998-09-17 | Leopold Kostal GmbH & Co KG, 58507 Lüdenscheid | Device for detecting influences affecting the visual quality for driving a motor vehicle |
| RU2416822C2 (en) * | 2006-03-21 | 2011-04-20 | Сейфгейт Интернэшнл Аб | Improved system for positioning aircraft on parking area |
| CN101936900A (en) * | 2010-06-12 | 2011-01-05 | 北京中科卓视科技有限责任公司 | Video-based visibility detecting system |
| CN102262092B (en) * | 2011-04-29 | 2013-03-20 | 中国气象局北京城市气象研究所 | Visibility measurement system and method |
| CN104777103A (en) * | 2015-04-15 | 2015-07-15 | 西安灏通节能工程设备有限公司 | Sight distance visibility meter and measuring method thereof |
-
2016
- 2016-03-29 CN CN201610185869.7A patent/CN105894500B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN105894500A (en) | 2016-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Koretsky et al. | Tutorial on Electro-Optical/Infrared (EO/IR) Theory and Systems | |
| CN102853916B (en) | Method and system for conducting remote infrared temperature measurement on coal pile surfaces | |
| CN102175613B (en) | Image-brightness-characteristic-based pan/tilt/zoom (PTZ) video visibility detection method | |
| CN102262092B (en) | Visibility measurement system and method | |
| CN105910712B (en) | Five-channel self-adaption two-dimensional temperature field measuring apparatus and its measurement method | |
| KR101735557B1 (en) | System and Method for Collecting Traffic Information Using Real time Object Detection | |
| GB2471952A (en) | Visual detection of clear air turbulence | |
| CN115507959B (en) | Infrared radiation characteristic analysis method for target detection | |
| CN104601953A (en) | Video image fusion-processing system | |
| Benes et al. | Multi-focus thermal image fusion | |
| CN113762161A (en) | Intelligent obstacle monitoring method and system | |
| Varjo et al. | Image based visibility estimation during day and night | |
| CN111898427A (en) | Multispectral pedestrian detection method based on feature fusion deep neural network | |
| CN105894500B (en) | A kind of visual range detection method based on image procossing | |
| CN113065277A (en) | Method for high-resolution remote sensing satellite flutter detection and modeling in cooperation with multi-load data | |
| CN108163223A (en) | A kind of portable aircraft infrared stealth capability evaluating device and method | |
| Portenier et al. | Cloud detection and visibility estimation during night time using thermal camera images | |
| Boettcher et al. | New target acquisition task for contemporary operating environments: personnel in MWIR, LWIR, and SWIR | |
| CN109781259B (en) | Method for accurately measuring infrared spectrum of small aerial moving target through spectrum correlation | |
| Chen et al. | A new algorithm for calculating the daytime visibility based on the color digital camera | |
| Wang et al. | Vehicle Micro-Trajectory Automatic Acquisition Method Based on Multi-Sensor Fusion | |
| CN112437216A (en) | Image optimization method and device | |
| Mostafa et al. | Determination of the cloud coverage using ground based camera images in the visible and infrared spectral range | |
| Starr | Rangefinding in fire smoke environments | |
| CN111504912A (en) | Air pollution observation system and method based on image recognition |
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 |