CN102044152A - Day and night video detecting method and device - Google Patents
Day and night video detecting method and device Download PDFInfo
- Publication number
- CN102044152A CN102044152A CN2010105538031A CN201010553803A CN102044152A CN 102044152 A CN102044152 A CN 102044152A CN 2010105538031 A CN2010105538031 A CN 2010105538031A CN 201010553803 A CN201010553803 A CN 201010553803A CN 102044152 A CN102044152 A CN 102044152A
- Authority
- CN
- China
- Prior art keywords
- background image
- clock
- brightness
- smoothing processing
- round
- 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.)
- Granted
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a day and night video detecting method and a day and night video detecting device. The method comprises: extracting a background image from a preset reference area; recording the brightness average value of the background image in a statistic period; performing smooth treatment on the recorded brightness average values of a plurality of continuous statistic periods; and determining the change tendency of the brightness average values of the background image according to the recorded brightness average value data of the plurality of continuous statistic periods. The invention also discloses a day and night video detecting device. The method and the device can reduce the probability that the feature data in the reference region are interfered and improve the stability of a detection result.
Description
Technical field
The present invention relates to intelligent transportation monitoring technique field, particularly a kind of video detecting method round the clock and device.
Background technology
Intelligent transportation system (ITS) is as a kind of advanced technology, apply to whole traffic management system effectively, and obtain the attention of various countries, can provide transport information in time, accurately for traffic department, thereby make traffic management control system effectively adapt to various traffics, utilization various control system, rationally dredge or allocate transport power at macroscopical relatively height, thereby the performance traffic control system of maximum efficiency is in the accuracy and the control of aspects such as traffic monitoring, traffic control, access and exit control, rescue management.
In intelligent transportation system, traffic monitoring is the intelligentized prerequisite of traffic administration, and setting up the traffic automatic monitored control system also just becomes the intelligentized top priority of traffic administration.The main target of traffic automatic monitored control system is to obtain road information and vehicle behavioural information, it is traffic events, comprising vehicle flowrate, the speed of a motor vehicle, following distance, type of vehicle, roadway occupancy, vehicle illegal information, traffic hazard detection, road meteorology, road construction, video monitor image etc., mainly lay particular emphasis on to the macro adjustments and controls of road with to the improvement of road act of violating regulations, transport development planning is had crucial meaning.
For the video analysis in the traffic monitoring, because the difference of day and night light, road lighting is widely different, need to adopt different detection methods to finish to vehicle detection, and then finish transport information and detect automatically, the handoff procedure of the vehicle detecting algorithm on this night and daytime is called conversion round the clock, and the core of conversion is detection and judgement to round the clock round the clock, i.e. detection method round the clock.
Existing detection method round the clock has following several:
1, directly detect round the clock by the time, this method defective clearly can't adapt to overcast and rainy influence to illumination on the one hand, can not adapt to the variation of season and region on the other hand.
2, detect feature object state in the camera views and judge round the clock and change, street lamp etc. for example, but on different roads, being difficult to search out unified standard, adaptability is poor.
3, the change-detection of utilizing light sensitive component to survey illumination changes round the clock, and this method is more effective, but photo-sensitive cell is installed cost is increased, and the life-span of photo-sensitive cell is limited, the installation site is positioned at outside the rack, and is destroyed easily, also increased the burden and the cost of later maintenance.
4, utilize the brightness of video image to detect round the clock, also claim Video Detection round the clock.This method is utilized some characteristic area in the image of shot by camera to analyze light to change, thereby realizes detection round the clock.
In the present switch technology round the clock, consider the balance of cost, maintenance and detection effect, generally use the 4th kind of detection method round the clock, i.e. Video Detection method round the clock, this method cost is low, do not need extra cost and maintenance, and applicability is extensive.
In Video Detection method round the clock, the key issue that influences the final detection result correctness has two, and the one, the selection problem in image reference zone, the 2nd, for the specific analytical method problem of image in the reference zone.
Selection for the image reference zone, mainly contain two kinds on sky and road surface, for selecting the method that day dummy section is analyzed in the image, because this method requires in the monitored picture sky picture to be arranged, and in the present traffic surveillance and control system, overwhelming majority cameras all are fixing seats in the plane, must have the sky picture to be difficult to accomplish that for the camera of these fixed angles adaptability is poor.Therefore, the zone conduct of general selection road surface is with reference to the zone in the existing video image analysis method, yet, no matter be to select sky or zone, road surface to analyze, the characteristic that obtains from reference zone is as monochrome information etc., all may be subjected to the influence of some disturbing factors in the image, for example the zone, road surface can be subjected to the influence of vehicle at night light, and day dummy section can be subjected to the influence of cloud layer, sand and dust etc., makes testing result be affected.
Specific analytical method for image in the reference zone, at present generally be utilize mean flow rate value directly rule of thumb value judge, this method real-time is good, do not need statistical study, but stability is not good enough, is easy to be subjected to the weather of variation suddenly and the influence of light, as car light, black clouds etc., cause the testing result erroneous judgement, thereby can not guarantee correctly to carry out the switching of vehicle detecting algorithm.
Summary of the invention
The embodiment of the invention provides a kind of detection method round the clock, can reduce the disturbed possibility of characteristic of reference zone, improves the stability of testing result.
The embodiment of the invention provides a kind of pick-up unit round the clock, can reduce the disturbed possibility of brightness data of reference zone, improves the stability of testing result.
For achieving the above object, technical scheme of the present invention specifically is achieved in that
A kind of video detecting method round the clock, this method comprises:
From the reference zone that sets in advance, extract background image;
The brightness average of described background image in the record measurement period;
Brightness mean data to continuous a plurality of measurement periods of described record is carried out smoothing processing;
Determine the variation tendency of the brightness average of described background image according to the brightness mean data of the measurement period of described continuous a plurality of records.
Preferably, the brightness average of described background image in the described record measurement period comprises:
When the brightness average of described background image surpasses or is lower than default thresholding, the step of background image brightness average in the executive logging measurement period.
Preferably, described smoothing processing comprises:
Adopt median filtering method that the brightness mean data of a plurality of measurement periods is handled.
Preferably, the brightness mean data of described measurement period according to a plurality of records is determined to comprise the variation tendency of the brightness average of background image:
Determine the variation tendency of this smoothing processing background luminance average in the cycle with the smoothing processing ratio that various Grad occur in the cycle;
Described gradient is: the difference between the data of two measurement periods of predetermined interval.
Preferably, describedly from the reference zone that sets in advance, extract background image, comprising:
Employing separates the movable information in the scene, thereby extracts the difference image method of static background image, or the Target Recognition method, or other motion detection algorithms.
Preferably, this method further comprises:
One day time is divided into dawn, daytime, dusk and 4 stages of night, in the stage at dawn, if the result of trend analysis be from steadily to rising, then vehicle detecting algorithm is switched to the algorithm on daytime; In the stage at dusk, if the result of trend analysis be from steadily to decline, then vehicle detecting algorithm is switched to the algorithm at night.
A kind of video detecting device round the clock, this device comprises:
The background extracting module is used for extracting background image from the reference zone that sets in advance;
Data recordin module links to each other with described background extracting module, is used to write down the brightness average of described background image in the measurement period;
The smoothing processing module links to each other with described data recordin module, is used for the brightness mean data of continuous a plurality of measurement periods of described record is carried out smoothing processing;
The trend judge module links to each other with described smoothing processing module, is used for determining according to the brightness mean data of the measurement period of described continuous a plurality of records the variation tendency of the brightness average of described background image.
Preferably, described data recordin module comprises:
Threshold decision unit, the brightness average that is used to judge described background image surpass or are lower than default thresholding;
Record cell is used for when the brightness average that described threshold decision unit judges goes out background image surpasses default thresholding background image brightness average in the record measurement period; When perhaps the brightness average that goes out background image in described threshold decision unit judges is lower than default thresholding, background image brightness average in the record measurement period.
Preferably, described smoothing processing module adopts median filtering method that the brightness mean data of a plurality of measurement periods is handled.
Preferably, described trend judge module comprises:
The gradient calculation unit is used for the difference between the data of two measurement periods of predetermined interval calculating;
The trend judging unit, the smoothing processing ratio that various Grad occur in the cycle that is used for calculating according to described gradient calculation unit is determined the variation tendency of this smoothing processing cycle background luminance average.
Preferably, described background extracting module adopts separates the movable information in the scene, thereby extracts the difference image method of static background image, or the Target Recognition method, or other motion detection algorithms extract background image from the reference zone that sets in advance.
Preferably, this device further comprises:
The algorithm handover module, the time that is used for one day is divided into dawn, daytime, dusk and 4 stages of night, in the stage at dawn, if the analysis result that obtains of described trend judge module for from steadily to rising, then vehicle detecting algorithm is switched to the algorithm on daytime; In the stage at dusk, if the analysis result that obtains of described trend judge module for from steadily to decline, then vehicle detecting algorithm is switched to the algorithm at night.
As seen from the above technical solutions, this video detecting method round the clock of the present invention and device, by extracting multiple modes such as background image, gradient data, trend analysis, reduced the disturbed possibility of characteristic of reference zone, improved the stability of testing result.
Description of drawings
Fig. 1 is the process flow diagram of video detecting method round the clock of the embodiment of the invention;
Fig. 2 is the structural representation of video detecting device round the clock of the embodiment of the invention;
Fig. 3 is the data recordin module structural representation of the embodiment of the invention;
Fig. 4 is the trend judge module structural representation of the embodiment of the invention;
Fig. 5 is the structural representation of video detecting device round the clock of another embodiment of the present invention.
Embodiment
For making purpose of the present invention, technical scheme and advantage clearer, below with reference to the accompanying drawing embodiment that develops simultaneously, the present invention is described in more detail.
The present invention mainly is when obtaining the characteristic of reference zone, the background image of reference zone is partly separated, thereby obtain the characteristic in the background image of reference zone, thereby can reduce the interference of vehicle light etc. in the reference zone prospect, and in the concrete analysis process, then adopted the gradient analysis method of original creation, by periodicity, in real time the gradient of the brightness data of reference zone is carried out statistical study, the general trend that analyzes the reference zone brightness is for steady, rise and still descend, thereby the testing result of obtaining, because by statistics to cycle data trend, further reduced the interference of bursty data for actual conditions, so that the stability of testing result be further enhanced.
Fig. 1 is the process flow diagram of video detecting method round the clock of the embodiment of the invention, and as shown in Figure 1, this method comprises the steps:
Being provided with of reference zone can be all can in sky or the road surface of using always, in general, because it generally all is hard-wired being used for the traffic monitoring video camera, the scene variability is less, considers all traffic route monitoring, and road surface characteristic has unitarity, and its brightness is carved with different variations when different round the clock, preferred employing zone, road surface is a reference zone, and the setting of reference zone only need be enabled before carrying out video analysis, and the reference zone that is provided with is just adopted in check and analysis afterwards.If scene changes, then carry out road surface reference zone setting once more and get final product.
For the extraction of background image, can adopt the movable information in the scene is separated, thereby extract the difference image method of static background image, also can adopt other motion detection algorithms such as Target Recognition method to realize.Extract after the background image of reference zone, the analysis of follow-up all data all is based on this background image.
On traffic route, the variation of illumination can be directly reacted in the brightness of road surface reference zone, thus among this method embodiment with the brightness average of the background image of reference zone characteristic as subsequent analysis.
Though use the background image of reference zone can suppress the influence of most car light light, but halation for the long period existence, background image will be contaminated, still can impact testing result, in order to suppress this phenomenon, be measurement period with the regular hour in this step, in measurement period, the brightness average of each frame background image adds up, the brightness mean of mean of all background image frame is as the brightness average of background image in this measurement period in fetch cycle, the brightness average of background image in each measurement period is noted, thus the influence of minimizing background contamination.
The concrete duration of measurement period can freely be set as required, is preferably 40 seconds.If video camera cuts out automatic gain, open high light inhibition etc., then the measurement period of video will shorten, because from the dusk to the night, or the time at night to dawn can shorten, 40 seconds be the normal empirical value of opening camera parameters, this value can be regulated and control according to camera parameters.
Step 103 is carried out smoothing processing to the brightness mean data of continuous a plurality of measurement periods;
The purpose of smoothing processing is in order better to remove the accidental data in the data, avoid the influence of emergency case to testing result, particularly, can adopt median filtering method in this step, it is a kind of basic skills of data processing, statistics is got intermediate value, can carry out smoothly data, remove burr, brightness mean data in continuous a plurality of measurement periods is filtered, and these a plurality of measurement periods can be described as the smoothing processing cycle, and concrete duration can freely be set as required, preferably 30 measurement period data are carried out smoothing processing, promptly the smoothing processing cycle is 30 measurement periods.
In addition, in order to guarantee real-time, when carrying out smoothing processing, if have new measurement period data value to add, then immediately statistics upgraded, promptly, each measurement period all will upgrade the statistics chained list one time, for follow-up analysis of trend provides real-time Data Source.
Step 104 is determined the variation tendency of background image brightness average.
Need in this step the background image brightness mean data after the smoothing processing is carried out trend analysis, analyze and write down these data show as dullly rise, dull decline or trend stably, for the switching of final traffic detection method provides foundation.As switching foundation, the inhibiting effect that can utilize accumulative total alternation efficient to disturb for data further strengthens the stability of testing result with trend.
The concrete time of trend analysis is in the time of can being chosen in each smoothing processing end cycle, concrete grammar can adopt the gradient analysis method, so-called Grad, it is the difference between the data of two measurement periods of certain intervals, can be on the occasion of (data of the measurement period that the data of the measurement period after the time leans on are more forward than the time are big), also can be negative value (with on the occasion of opposite), also can be zero.In the embodiment of the invention, can determine the variation tendency of this smoothing processing background luminance average in the cycle with the smoothing processing ratio that various Grad occur in the cycle, with a smoothing processing cycle be 30 measurement periods, the gradient of data is an example at interval, 30 brightness mean data have 28 Grad (if two data computation Grad in interval, then have 27 Grad, by that analogy), if the ratio of negative value is more than or equal to 80% in these Grad, then determine variation tendency for descending, otherwise, on the occasion of ratio more than or equal to 80%, determine that then variation tendency for rising, is defined as under other situation steadily.Certainly, the measurement period number that data break when ratio, compute gradient appear in the Grad that specifically is used for determining trend and smoothing processing cycle comprise can be adjusted as required.
Determine the brightness Change in Mean trend of background image by above-mentioned steps after, can change round the clock according to this variation tendency, if the result of trend analysis descends, the intensity of illumination that the road surface then is described is descending, and weather just enters night from daytime, otherwise illustrates that then intensity of illumination rises, weather just enters daytime from night, if the result of trend analysis is steady, not too big variation of intensity of illumination then is described, weather does not change.After obtaining the result of trend analysis, can carry out the switching of follow-up vehicle detecting algorithm according to this result.
In addition, in step 102 record measurement period before the brightness average of background image, can add a trigger condition, the empirical value of a brightness average promptly is set according to image scene, for example be set to 60, can be when the brightness average of background image surpasses or is lower than this empirical value, just trigger the process of the record brightness mean data of step 102, otherwise do not write down this brightness mean data, do like this and can further reduce this required data processing quantity of detection method round the clock, reduce the waste of handling resource.Particularly, this experience thresholding can be carried out up-down adjustment according to front end camera characteristics and scene, and scope can be between [20,90].
At last, after the testing result that is detected round the clock, the switching of carrying out follow-up vehicle detecting algorithm according to the result who detects has a variety of implementation methods, for example one day time can be divided into dawn, daytime, dusk and 4 stages of night, transformational analysis round the clock can only be carried out at dawn and stage at dusk, rather than round-the-clock carrying out, thereby further reduces calculated amount.During concrete the switching can: in the stage at dawn, if the result of trend analysis be from steadily to rising, then switch the night algorithm to the algorithm on daytime; In the stage at dusk, if continuously the result of trend analysis be from steadily to decline, then from daytime algorithm pattern switch to the algorithm pattern at night.In addition, also can be in the stage at dawn, if the result of trend analysis is from steadily to rising to steadily, then switching the night algorithm to the algorithm on daytime again; In the stage at dusk, if the result of trend analysis from steadily to descending again to steadily, then from daytime algorithm switch to the algorithm at night.Concrete grammar much can not be enumerated one by one, repeats no more here.
Fig. 2 is the structural representation of pick-up unit round the clock of the embodiment of the invention, and as shown in Figure 2, this device comprises:
Described background extracting module 201 can adopt separates the movable information in the scene, thereby extracts the difference image method of static background image, or the Target Recognition method, or other motion detection algorithms extract background image from the reference zone that sets in advance.
Described smoothing processing module 203 can adopt median filtering method that the brightness mean data of a plurality of measurement periods is handled.
Wherein, described data recordin module 202 can comprise as shown in Figure 3:
In addition, as shown in Figure 4, described trend judge module 204 can comprise:
In addition, as shown in Figure 5, remove background extracting module 201, data recordin module 202, smoothing processing module 203, outside the trend judge module 204, this device can further include:
Algorithm handover module 501, link to each other with described trend judge module 204, the time that is used for one day is divided into dawn, daytime, dusk and 4 stages of night, in the stage at dawn, if the analysis result that obtains of described trend judge module 204 for from steadily to rising, then vehicle detecting algorithm is switched to the algorithm on daytime; In the stage at dusk, if the analysis result that obtains of described trend judge module for from steadily to decline, then vehicle detecting algorithm is switched to the algorithm at night.
By the above embodiments as seen, this video detecting method round the clock of the present invention and device are partly separated the background image of reference zone, thereby obtain the characteristic in the background image of reference zone, reduced the interference of vehicle light etc. in the reference zone prospect, and in the concrete analysis process, then adopted the gradient analysis method of original creation, by periodicity, in real time the gradient of the brightness data of reference zone is carried out statistical study, the general trend that analyzes the reference zone brightness is for steady, rise and still descend, thereby the testing result of obtaining, because by statistics to cycle data trend, further reduced the interference of bursty data for actual conditions, so that the stability of testing result be further enhanced.
Claims (12)
1. video detecting method round the clock is characterized in that this method comprises:
From the reference zone that sets in advance, extract background image;
The brightness average of described background image in the record measurement period;
Brightness mean data to continuous a plurality of measurement periods of described record is carried out smoothing processing;
Determine the variation tendency of the brightness average of described background image according to the brightness mean data of the measurement period of the continuous a plurality of records after the described smoothing processing.
2. video detecting method round the clock as claimed in claim 1 is characterized in that, the brightness average of described background image in the described record measurement period comprises:
When the brightness average of described background image surpasses or is lower than preset threshold value, the step of background image brightness average in the executive logging measurement period.
3. video detecting method round the clock as claimed in claim 1 is characterized in that, described smoothing processing comprises:
Adopt median filtering method that the brightness mean data of a plurality of measurement periods is handled.
4. video detecting method round the clock as claimed in claim 1 is characterized in that, the brightness mean data of described measurement period according to a plurality of records is determined to comprise the variation tendency of the brightness average of background image:
Determine the variation tendency of this smoothing processing background luminance average in the cycle with the smoothing processing ratio that various Grad occur in the cycle;
Described Grad is: the difference between the data of two measurement periods of predetermined interval.
5. video detecting method round the clock as claimed in claim 1 is characterized in that, describedly extracts background image from the reference zone that sets in advance, and comprising:
Employing separates the movable information in the scene, thereby extracts the difference image method of static background image, or the Target Recognition method, or other motion detection algorithms.
6. video detecting method round the clock as claimed in claim 1 is characterized in that, this method further comprises:
One day time is divided into dawn, daytime, dusk and 4 stages of night, in the stage at dawn, if the result of trend analysis be from steadily to rising, then vehicle detecting algorithm is switched to the algorithm on daytime; In the stage at dusk, if the result of trend analysis be from steadily to decline, then vehicle detecting algorithm is switched to the algorithm at night.
7. video detecting device round the clock is characterized in that this device comprises:
The background extracting module is used for extracting background image from the reference zone that sets in advance;
Data recordin module links to each other with described background extracting module, is used to write down the brightness average of described background image in the measurement period;
The smoothing processing module links to each other with described data recordin module, is used for the brightness mean data of continuous a plurality of measurement periods of described record is carried out smoothing processing;
The trend judge module links to each other with described smoothing processing module, is used for determining according to the brightness mean data of the measurement period of the continuous a plurality of records after the described smoothing processing variation tendency of the brightness average of described background image.
8. video detecting device round the clock as claimed in claim 1 is characterized in that, described data recordin module comprises:
Threshold decision unit, the brightness average that is used to judge described background image surpass or are lower than preset threshold value;
Record cell is used for when the brightness average that described threshold decision unit judges goes out background image surpasses preset threshold value background image brightness average in the record measurement period; When perhaps the brightness average that goes out background image in described threshold decision unit judges is lower than preset threshold value, background image brightness average in the record measurement period.
9. video detecting device round the clock as claimed in claim 1 is characterized in that, described smoothing processing module adopts median filtering method that the brightness mean data of a plurality of measurement periods is handled.
10. video detecting device round the clock as claimed in claim 1 is characterized in that, described trend judge module comprises:
The gradient calculation unit is used for the difference between the data of two measurement periods of predetermined interval calculating;
The trend judging unit, the smoothing processing ratio that various Grad occur in the cycle that is used for calculating according to described gradient calculation unit is determined the variation tendency of this smoothing processing cycle background luminance average.
11. video detecting device round the clock as claimed in claim 1, it is characterized in that, described background extracting module adopts separates the movable information in the scene, thereby extract the difference image method of static background image, or the Target Recognition method, or other motion detection algorithms extract background image from the reference zone that sets in advance.
12. video detecting device round the clock as claimed in claim 1 is characterized in that, this device further comprises:
The algorithm handover module, the time that is used for one day is divided into dawn, daytime, dusk and 4 stages of night, in the stage at dawn, if the analysis result that obtains of described trend judge module for from steadily to rising, then vehicle detecting algorithm is switched to the algorithm on daytime; In the stage at dusk, if the analysis result that obtains of described trend judge module for from steadily to decline, then vehicle detecting algorithm is switched to the algorithm at night.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010553803 CN102044152B (en) | 2010-11-19 | 2010-11-19 | Day and night video detecting method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010553803 CN102044152B (en) | 2010-11-19 | 2010-11-19 | Day and night video detecting method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102044152A true CN102044152A (en) | 2011-05-04 |
| CN102044152B CN102044152B (en) | 2013-10-30 |
Family
ID=43910256
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010553803 Active CN102044152B (en) | 2010-11-19 | 2010-11-19 | Day and night video detecting method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102044152B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104200473A (en) * | 2014-09-02 | 2014-12-10 | 中航高科技发展有限公司 | Image background subtraction based dynamic vision measuring method |
| CN105187731A (en) * | 2015-09-18 | 2015-12-23 | 中电海康集团有限公司 | Day-night switching method and device based on forward collision warning |
| CN105979153A (en) * | 2016-06-28 | 2016-09-28 | 深圳市新国都支付技术有限公司 | Day and night mode identification and switching method and system |
| CN103702038B (en) * | 2013-12-31 | 2017-04-19 | 浙江宇视科技有限公司 | Dynamic threshold value-based method and device for performing automatic day and night switching |
| CN109155107A (en) * | 2016-03-22 | 2019-01-04 | 德尔福技术有限公司 | Sensory perceptual system for automated vehicle scene perception |
| CN117252449A (en) * | 2023-11-20 | 2023-12-19 | 水润天府新材料有限公司 | Full-penetration drainage low-noise pavement construction process and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101034005A (en) * | 2007-04-12 | 2007-09-12 | 上海交通大学 | Switching method between videos for detecting vehicles during day-night |
| CN201274542Y (en) * | 2008-09-01 | 2009-07-15 | 安徽科力信息产业有限责任公司 | Video monitoring apparatus for day and night |
| CN101556739A (en) * | 2009-05-14 | 2009-10-14 | 浙江大学 | Vehicle detecting algorithm based on intrinsic image decomposition |
| CN101770571A (en) * | 2009-12-29 | 2010-07-07 | 北京世纪高通科技有限公司 | Method and device for detecting vehicle at night |
| CN101872526A (en) * | 2010-06-01 | 2010-10-27 | 重庆市海普软件产业有限公司 | Smoke and fire intelligent identification method based on programmable photographing technology |
-
2010
- 2010-11-19 CN CN 201010553803 patent/CN102044152B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101034005A (en) * | 2007-04-12 | 2007-09-12 | 上海交通大学 | Switching method between videos for detecting vehicles during day-night |
| CN201274542Y (en) * | 2008-09-01 | 2009-07-15 | 安徽科力信息产业有限责任公司 | Video monitoring apparatus for day and night |
| CN101556739A (en) * | 2009-05-14 | 2009-10-14 | 浙江大学 | Vehicle detecting algorithm based on intrinsic image decomposition |
| CN101770571A (en) * | 2009-12-29 | 2010-07-07 | 北京世纪高通科技有限公司 | Method and device for detecting vehicle at night |
| CN101872526A (en) * | 2010-06-01 | 2010-10-27 | 重庆市海普软件产业有限公司 | Smoke and fire intelligent identification method based on programmable photographing technology |
Non-Patent Citations (1)
| Title |
|---|
| 吴海涛: "复杂环境下的夜间车辆检测与交通视频昼夜亮度变化模型研究方法", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 06, 15 June 2007 (2007-06-15), pages 140 - 307 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103702038B (en) * | 2013-12-31 | 2017-04-19 | 浙江宇视科技有限公司 | Dynamic threshold value-based method and device for performing automatic day and night switching |
| CN104200473A (en) * | 2014-09-02 | 2014-12-10 | 中航高科技发展有限公司 | Image background subtraction based dynamic vision measuring method |
| CN105187731A (en) * | 2015-09-18 | 2015-12-23 | 中电海康集团有限公司 | Day-night switching method and device based on forward collision warning |
| CN105187731B (en) * | 2015-09-18 | 2018-03-16 | 中电海康集团有限公司 | A kind of switching method round the clock and device based on preceding anti-collision warning |
| CN109155107A (en) * | 2016-03-22 | 2019-01-04 | 德尔福技术有限公司 | Sensory perceptual system for automated vehicle scene perception |
| US11604474B2 (en) | 2016-03-22 | 2023-03-14 | Motional Ad Llc | Scenario aware perception system for an automated vehicle |
| CN105979153A (en) * | 2016-06-28 | 2016-09-28 | 深圳市新国都支付技术有限公司 | Day and night mode identification and switching method and system |
| CN117252449A (en) * | 2023-11-20 | 2023-12-19 | 水润天府新材料有限公司 | Full-penetration drainage low-noise pavement construction process and system |
| CN117252449B (en) * | 2023-11-20 | 2024-01-30 | 水润天府新材料有限公司 | Full-penetration drainage low-noise pavement construction process and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102044152B (en) | 2013-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102044152B (en) | Day and night video detecting method and device | |
| KR101095528B1 (en) | An outomatic sensing system for traffic accident and method thereof | |
| CN102231236B (en) | Method and device for counting vehicles | |
| CN103069434B (en) | For the method and system of multi-mode video case index | |
| KR101392294B1 (en) | Video segmentation using statistical pixel modeling | |
| CN111429726B (en) | Monitoring video illegal parking vehicle detection and management method and corresponding system | |
| KR100834550B1 (en) | Detecting method at automatic police enforcement system of illegal-stopping and parking vehicle and system thereof | |
| CN105654508B (en) | Monitor video method for tracking moving target and system based on adaptive background segmentation | |
| CN104506804A (en) | Device and method for monitoring abnormal behavior of motor vehicle on expressway | |
| CN113256965A (en) | Traffic risk sensing system and method based on smart city | |
| CN105678730A (en) | Camera movement self-detecting method on the basis of image identification | |
| CN111163285A (en) | High-altitude falling object monitoring method and system and computer readable storage medium | |
| CN112112629A (en) | Safety business management system and method in drilling operation process | |
| CN105654734A (en) | Data processing method and device for vehicle violations | |
| CN106781460B (en) | A kind of road section traffic volume state determines method and device | |
| CN104464305A (en) | Intelligent vehicle converse driving detecting device and method | |
| CN104200466A (en) | Early warning method and camera | |
| CN104601965A (en) | Camera shielding detection method | |
| CN111523397A (en) | Intelligent lamp pole visual identification device, method and system and electronic equipment | |
| CN106710253B (en) | High-reliability intelligent intersection traffic control system and control method | |
| JP7305965B2 (en) | Video surveillance system parameter setting method, device and video surveillance system | |
| CN105046948A (en) | System and method of monitoring illegal traffic parking in yellow grid line area | |
| CN116580547A (en) | Intelligent control method and system for buried traffic signal lamp | |
| CN102542673A (en) | Automatic teller machine (ATM) pre-warning method and system based on computer vision | |
| CN117409588A (en) | Traffic abnormality monitoring method and system based on geomagnetic video data fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |