CN113554872A - Detection early warning method and system for traffic intersection and curve - Google Patents
Detection early warning method and system for traffic intersection and curve Download PDFInfo
- Publication number
- CN113554872A CN113554872A CN202110814416.7A CN202110814416A CN113554872A CN 113554872 A CN113554872 A CN 113554872A CN 202110814416 A CN202110814416 A CN 202110814416A CN 113554872 A CN113554872 A CN 113554872A
- Authority
- CN
- China
- Prior art keywords
- image
- target
- curve
- intersection
- detection
- 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
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/048—Detecting movement of traffic to be counted or controlled with provision for compensation of environmental or other condition, e.g. snow, vehicle stopped at detector
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a detection early warning method and a system for a traffic intersection and a curve, which continuously acquire original images at a first preset distance from a target intersection or a target curve by using camera equipment; defogging the original image through a defogging algorithm for generating a countermeasure network based on conditions to obtain a fog-free image; detecting vehicles and pedestrians in the fog-free image through an image detection network; and when the target object is detected, displaying the target object and the distance between the target object and the intersection or the curve through warning equipment. The system can monitor complex road sections such as curves and intersections in real time, provide accurate and timely road condition information for drivers, fill up the blank of the rural curve and intersection early warning scheme, is suitable for complex rural environments, and greatly reduces the occurrence rate of rural traffic accidents.
Description
Technical Field
The invention relates to the technical field of image detection, in particular to a detection early warning method and a detection early warning system for traffic intersections and curves.
Background
In rural traffic in China, a large number of traffic accidents occur every year, and a lot of people are injured, disabled and even killed. There are many reasons for traffic accidents in rural traffic, and one of the important reasons is that the road condition of rural roads is complex, and drivers cannot timely acquire traffic condition information, especially on roads where the road condition is not easy to see clearly, such as curves or intersections, thereby causing various traffic accidents. Therefore, an advanced detection technology based on a camera is introduced to monitor complex road sections such as curves and intersections in real time, and the road condition information is extracted and displayed, so that accurate and real-time road condition information is provided for drivers of the current road sections, and the effect of greatly reducing rural traffic accidents is achieved. However, some current detection technologies are not well applicable to rural traffic, and due to the fact that rural roads are close to geographical environments such as mountainous regions and farmlands, wind and sand are large, fog is large, and most of snap photos for detection are not clear.
The application scenes of the detection algorithm in the prior art are all scenes with good environment and clear snapshot pictures, and the detection algorithm is not friendly to further complicated and poor environment. The system is not introduced into practical use in rural environment, and part of the reason is that the rural environment is complex and good working effect is difficult to realize. The current popular target detection technology based on deep learning is based on the detection of pictures shot by a camera, so that the requirement on the quality of input images is high, and if the input pictures have noise disturbance, the detection result is greatly influenced.
Disclosure of Invention
In order to overcome the technical defects, the invention aims to provide a curve which is suitable for curves with larger environmental interference factors such as fog, wind sand and the like.
The invention discloses a detection and early warning method for traffic intersections and curves, which comprises the following steps: continuously acquiring original images at a first preset distance from a target intersection or a target curve by using camera equipment, wherein the camera equipment is arranged at an entrance of the target intersection or at the outer side of the target curve; defogging the original image through a defogging algorithm of a countermeasure network generated based on conditions, wherein the countermeasure network generated based on conditions comprises a generator and a discriminator; inputting the original image into a generator to obtain a fog-free image; comparing the fog-free image with the original image through a discriminator, and judging the similarity through a twin neural network of the discriminator to obtain whether the fog-free image is the fog-free image of the original image, if so, outputting the fog-free image; detecting a target object in the fog-free image through an image detection network, wherein the target object comprises a vehicle and a pedestrian; when the target object is detected, displaying the distance between the target object and the intersection or the curve through warning equipment; the warning equipment is arranged at a second preset distance away from the target intersection or the target curve and is positioned at the inner side of the target intersection or the target curve.
Preferably, the loss function of the defogging algorithm based on the condition generation countermeasure network is as follows:
LossG＝log(D(x,G(x)))*W1+(y-G(x))2*W2(ii) a The penalty function of the discriminator is:
LossG ═ - (log (D (x, y)) + log (1-D (x, g (x))); wherein G is a generator, D is a discriminator, x is an original image, and y is a fog-free image.
Preferably, the generator is a Unet + + network, a down-sampling channel and four up-sampling channels are adopted, and a connection structure is utilized to perform feature superposition by a dimension splicing method; and in each convolution module in the down-sampling stage, a series-connected hole convolution group with a hole rate of [1,2,3,5] is used.
Preferably, the discriminator includes Block _1, Block _2, Block _3, and Block _4, where: block _2 is an acceptance-A module in an acceptance _ v4 network; block _3 is an acceptance-B module in an acceptance _ v4 network; block _4 is an acceptance-C module in an acceptance _ v4 network.
Preferably, when the target object is detected, displaying the target object and the distance between the target object and the intersection or the curve through the warning device further comprises: and displaying the lane where the target object is located through warning equipment.
Preferably, the method further comprises the following steps: taking a first preset time period as an interval, and storing a fog-free image with the highest definition in a plurality of fog-free images acquired within the first preset time as an analysis image; and if the traffic accident in the first preset time period is detected, acquiring the license plate number of the vehicle in the analysis image through image recognition.
Preferably, if a traffic accident within the first preset time period is detected, acquiring the license plate number of the vehicle in the analysis image through image recognition further includes: if the license plate number cannot be detected and obtained, the analysis image is marked and stored.
Preferably, the second predetermined distance is greater than 20 meters.
The invention also discloses a detection and early warning system positioned at the traffic intersection and the curve, which comprises a camera module, a processing module and a warning module; continuously acquiring original images at a first preset distance from a target intersection or a target curve by the camera module, wherein the camera module is arranged at an entrance of the target intersection or at the outer side of the target curve; defogging the original image through a defogging algorithm of a countermeasure network generated based on conditions in the processing module to obtain a fog-free image; detecting a target object in the fog-free image through an image detection network in the processing module, wherein the target object comprises a vehicle and a pedestrian; when the target object is detected, the distance between the target object and the intersection or the curve is displayed through the warning module; the warning module is arranged at a second preset distance from the target intersection or the target curve and is positioned on the inner side of the target curve.
Preferably, the camera module comprises a camera provided with a holder; the warning module comprises a large screen display provided with a lamp and a loudspeaker.
After the technical scheme is adopted, compared with the prior art, the method has the following beneficial effects:
1. the photographed road condition pictures are subjected to defogging treatment through the Unet + + network structure, so that interference factors in the images are reduced, and the method is suitable for rural curves, intersections and the like with complicated road environments such as large dust and sand, heavy fog and the like; color distortion hardly occurs, a fog-free scene is highly restored, the defogging accuracy is improved, images with higher definition are input for a detection process, and the precision of various detection algorithms in a rural environment with a complex road environment is ensured; the system can monitor complex road sections such as curves and intersections in real time, provide accurate and timely road condition information for drivers, fill up the blank of the rural curve and intersection early warning scheme, is suitable for complex rural environments, and greatly reduces the occurrence rate of rural traffic accidents.
Drawings
FIG. 1 is a flow chart of a detection and early warning method provided by the present invention;
FIG. 2 is a flow chart of a defogging algorithm for generating a countermeasure network based on conditions according to the present invention;
fig. 3 is a structure diagram of a net + + network according to the present invention;
FIG. 4 is a block diagram of a discriminator network according to the present invention;
FIG. 5 is a Block _1 Block diagram of the arbiter network of FIG. 4 according to the present invention;
fig. 6 is a structural diagram of a twin neural network of the discriminator network provided by the present invention.
Detailed Description
The advantages of the invention are further illustrated in the following description of specific embodiments in conjunction with the accompanying drawings.
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.
The terminology used in the present disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. As used in this disclosure and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It is to be understood that although the terms first, second, third, etc. may be used herein to describe various information, such information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present disclosure. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.
In the description of the present invention, it is to be understood that the terms "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on those shown in the drawings, and are used merely for convenience of description and for simplicity of description, and do not indicate or imply that the referenced devices or elements must have a particular orientation, be constructed in a particular orientation, and be operated, and thus, are not to be construed as limiting the present invention.
In the description of the present invention, unless otherwise specified and limited, it is to be noted that the terms "mounted," "connected," and "connected" are to be interpreted broadly, and may be, for example, a mechanical connection or an electrical connection, a communication between two elements, a direct connection, or an indirect connection via an intermediate medium, and specific meanings of the terms may be understood by those skilled in the art according to specific situations.
In the following description, suffixes such as "module", "component", or "unit" used to denote elements are used only for facilitating the explanation of the present invention, and have no specific meaning in themselves. Thus, "module" and "component" may be used in a mixture.
Referring to the attached figure 1, the invention discloses a detection and early warning method for traffic intersections and curves, which comprises the following steps of real-time shooting, image defogging, image detection and intersection early warning:
s1, continuously acquiring original images at a first preset distance from a target intersection or a target curve through camera equipment;
s2, defogging the original image through a defogging algorithm of a countermeasure network generated based on the condition, wherein the countermeasure network generated based on the condition comprises a generator and a discriminator; inputting an original image into a generator to obtain a fog-free image; comparing the fog-free image with the original image through a discriminator, judging whether the fog-free image is the fog-free image of the original image or not through the similarity, and if so, outputting the fog-free image;
s3, detecting a target object in the fog-free image through an image detection network, wherein the target object comprises a vehicle and a pedestrian;
and S4, when the target object is detected, displaying the target object and the distance between the target object and the intersection or the curve through the warning equipment.
In the process of S1, in order to ensure the enough shooting range and improve the detection early warning effect, the shooting equipment is arranged at the entrance of the target intersection or at the outer side of the target curve. For example, for intersections, a camera is placed at each intersection entrance with the lens facing the inside of the road segment, and for tight curves, the camera is placed outside the curve.
For the warning device in S4, in order to ensure that the vehicle has enough braking distance after receiving the warning, the warning device should be located at a second preset distance from the target intersection or the target curve, and located at the inner side of the target intersection or the target curve. For example, for an intersection, the warning device apparatus is at least a second predetermined distance from the intersection; for a sharp curve section, the warning device is arranged on the inner side of the curve and at least a second preset distance away from the curve.
The warning content is road condition information in other road sections of the front intersection which cannot be seen on the current driving route, and comprises the following steps: whether there is a pedestrian, whether there is a vehicle, and the detected distance of the target object from the intersection.
In rural environments, wind sand and fog are large, shot pictures are fuzzy, and more noise disturbance exists, so that the detection precision is reduced rapidly, and therefore the input pictures need to be subjected to defogging treatment. The invention is realized by adopting the improvement of generating a countermeasure network (GAN) based on conditions, and the network structure is shown in figure 2.
The detection target mainly comprises pedestrian detection and vehicle detection, wherein the vehicle detection comprises the following types: passenger vehicle (including large and small passenger vehicles), agricultural vehicle (mainly tractors, rotary cultivators and tricycles), motorcycle, battery car, bicycle, etc. The detection algorithm adopts a Yolov3 improvement-based training test under rural scene data.
The loss function of the conditional generation-based anti-network defogging algorithm is: loss _ G ═ Loss _1 × W1+ Loss _2 × W2.
W1 and W2 are loss weights to equalize the capabilities of the discriminators and generators, preventing the discriminators from being too strong, returning an invalid gradient, or too weak, causing the network to converge poorly. W1 and W2 both default to 1 initially, and need to be adjusted according to the change situation of the loss value of the actual training when set in training. Wherein Loss _1 ═ log (D (x, g (x)); loss _2 is the Loss of the generated image g (x) and the fog-free image y, and due to the principle of fog distribution in the atmosphere, the noise in the fog-free image is relatively uniform, and there are almost no outliers, so the Loss of L2 with better robustness is adopted, and is defined as: loss _2 ═ (y-g (x)) 2. The resulting loss function of the generator is: LossG ═ log (D (x, g (x))) W1+(y-G(x))2*W2(ii) a Wherein G is a generator, D is a discriminator, x is an original image, y is a fog-free image, and z is a random noise vector. The generator generally adopts an encoder-decoder structure, and aims to input a foggy image and output a fogless image after passing through the generator.
In order to realize the removal of atomization with higher resolution at a pixel level, a generator is an Unnet + + network, referring to fig. 3, a down-sampling channel and four up-sampling channels are adopted, a DenseNet connection structure is utilized, feature superposition is carried out by a dimension splicing method, and different features of 4 levels are well utilized to realize fine-grained prediction. In the invention, in order to obtain a larger receptive field in the convolution process and fully utilize local and remote spatial information, a series cavity convolution group with a cavity rate of [1,2,3,5] is used in each convolution module in a lower sampling stage, so that the receptive field is increased, the defogging effect is improved, and the chessboard effect problem caused by cavity convolution is avoided.
It should be noted that the connection structure does not necessarily need to adopt a DenseNet connection structure, and connection structures of other networks may implement feature superposition.
The discriminator is used for comparing the similarity degree of the generated fog-free image and the target fog-free image, and adopts a twin neural network structure, wherein the network structure is as shown in figure 6 and comprises two networks with the same mode, namely the left side and the right side share network parameters; two samples are input, and two extracted feature vectors are output to compare the similarity degree of the two samples. Wherein, loss _ out is the Euclidean distance of the two extracted feature maps, and the loss function of the discriminator is as follows:
LossG＝-(log(D(x,y))+log(1-D(x,G(x))))。
the internal network structure of the arbiter is shown in fig. 4 and fig. 5, which includes Block _1, Block _2, Block _3, and Block _4, wherein: the structure of Block _1 is shown in FIG. 5, Block _2 is an acceptance-A module in an acceptance _ v4 network; block _3 is an acceptance-B module in an acceptance _ v4 network; block _4 is an acceptance-C module in an acceptance _ v4 network, adopts various sizes and separation convolution kernels, aims to comprehensively extract features from different-size visual fields, and sufficiently compares a fog-free image generated by a generator with an original fog-free image.
After the original fog-containing image is subjected to condition generation and confrontation network, the image after defogging can be generated, and the image can be input into the detection network for corresponding detection.
Preferably, in order to further improve the early warning effect, when the target object is detected, the displaying the distance between the target object and the intersection or the curve by the warning device further includes: the lane where the target object is located is displayed through the warning equipment, so that more comprehensive information early warning is given to vehicles coming and going, and driving safety is guaranteed.
The detection technology of the invention inputs images with higher definition for the detection process through the improved image defogging algorithm, ensures the precision of the detection algorithm in rural environment, fills the blank of road early warning scheme in rural traffic, and can effectively reduce the traffic accident rate of rural curves and complex intersections.
In addition, aiming at the phenomenon of hit-and-run in more traffic accidents in rural areas, the detection and early warning scheme also comprises a traffic event detection and license plate recognition algorithm. Taking a first preset time period as an interval, and storing a fog-free image with the highest definition in a plurality of fog-free images acquired within the first preset time as an analysis image; and if the traffic accident in the first preset time period is detected, acquiring the license plate number of the vehicle in the analysis image through image recognition.
Specifically, high-definition images are captured at certain intervals during detection, the images with the highest scores are captured and defogged from the images processed by the defogging algorithm and then are compared with the original images to be stored, and license plate recognition is performed on the captured images once traffic accidents at the intersection are detected.
If no traffic accident is detected within a certain time period from the appearance of the vehicle to the disappearance of the vehicle, the stored analysis picture is deleted, assuming that no analysis is necessary. Through the real-time snapshot detection, the phenomenon of causing and escaping after traffic accidents occur on rural roads can be effectively relieved.
The license plate recognition is performed on a vehicle with a license plate, the analysis image is marked and stored for a non-license vehicle such as a farm vehicle, and the analysis image is directly used as a basis for post-processing events in follow-up investigation.
Preferably, the second predetermined distance is greater than 20 meters.
Preferably, a plurality of warning devices can be continuously arranged on the same road, and even if the vehicle-driving personnel do not pay attention to the first warning device, the messages can be acquired through one or more following warning devices, so that the communication holes of the warning messages are prevented.
The warning devices are arranged at equal intervals, and can be arranged more densely closer to the target intersection or the target curve. For example, for the same intersection, three warning devices are arranged, and the distance between the warning device closest to the intersection and the second-closest warning device is smaller than the distance between the second-closest warning device and the last warning device.
Preferably, the display of the warning device closer to the target intersection or the target curve can be more greatly enclosed.
The invention designs and optimizes the detection method based on the curve and intersection scene which are most frequently subjected to traffic accidents in rural traffic, has high detection precision, can accurately extract road condition information of complex road sections in rural roads in real time, and displays the road condition information to drivers on the road sections, thereby reducing the occurrence rate of traffic accidents.
The invention also discloses a detection and early warning system for traffic intersections and curves, which comprises:
-a camera module providing real-time video and photo capture;
the processing module is used for carrying out real-time defogging processing and target object detection on the input video and the input photo to obtain a detection result;
the warning module displays the extracted detection result and provides real-time road condition information for the driver.
Specifically, the original images are continuously acquired at a first preset distance from the target intersection or the target curve by the camera module, and the camera module is arranged at an entrance of the target intersection or at the outer side of the target curve. Then, defogging processing is carried out on the original image through a defogging algorithm of a countermeasure network generated based on conditions in the processing module to obtain a fog-free image; detecting vehicles and pedestrians in the fog-free image through an image detection network in the processing module; when the target object is detected, the distance between the target object and the intersection or the curve is displayed through a warning module; the warning module is arranged at a second preset distance from the target intersection or the target curve and is positioned on the inner side of the target curve.
Better, camera module can guarantee wider camera shooting scope including the camera that is equipped with the cloud platform. The warning module comprises a large screen display provided with a lamp and a loudspeaker, and the driving is prompted by visual warning and sound warning.
For one, a plurality of warning devices are continuously arranged on the same road, and the warning devices are arranged at equal intervals, and can be arranged more densely closer to a target intersection or a target curve. The display of the warning equipment closer to the target intersection or the target curve can be enclosed to be larger, namely the warning words formed by the lamp are larger and the amplified playing sound is larger closer to the target intersection or the target curve.
It should be noted that the embodiments of the present invention have been described in terms of preferred embodiments, and not by way of limitation, and that those skilled in the art can make modifications and variations of the embodiments described above without departing from the spirit of the invention.
Claims (10)
1. A detection early warning method for traffic intersections and curves is characterized by comprising the following steps:
continuously acquiring original images at a first preset distance from a target intersection or a target curve by using camera equipment, wherein the camera equipment is arranged at an entrance of the target intersection or at the outer side of the target curve;
defogging the original image through a defogging algorithm of a countermeasure network generated based on conditions, wherein the countermeasure network generated based on conditions comprises a generator and a discriminator; inputting the original image into a generator to obtain a fog-free image; comparing the fog-free image with the original image through a discriminator, and judging the similarity through a twin neural network of the discriminator to obtain whether the fog-free image is the fog-free image of the original image, if so, outputting the fog-free image;
detecting a target object in the fog-free image through an image detection network, wherein the target object comprises a vehicle and a pedestrian;
when the target object is detected, displaying the distance between the target object and the intersection or the curve through warning equipment; the warning equipment is arranged at a second preset distance away from the target intersection or the target curve and is positioned at the inner side of the target intersection or the target curve.
2. The detection and warning method of claim 1, wherein the loss function of the conditional generation countermeasure network based defogging algorithm is:
LossG＝log(D(x,G(x)))*W1+(y-G(x))2*W2；
the penalty function of the discriminator is:
LossG＝-(log(D(x,y))+log(1-D(x,G(x))))；
wherein G is a generator, D is a discriminator, x is an original image, and y is a fog-free image.
3. The detection and early warning method according to claim 1, wherein the generator is a Unet + + network, one down-sampling channel and four up-sampling channels are adopted, and a connection structure is utilized to perform feature superposition by a dimension splicing method;
and in each convolution module in the down-sampling stage, a series-connected hole convolution group with a hole rate of [1,2,3,5] is used.
4. The detection and warning method according to claim 1, wherein the discriminator includes Block _1, Block _2, Block _3, and Block _4, and wherein: block _2 is an acceptance-A module in an acceptance _ v4 network; block _3 is an acceptance-B module in an acceptance _ v4 network; block _4 is an acceptance-C module in an acceptance _ v4 network.
5. The detection and warning method of claim 1, wherein when the target object is detected, the step of displaying the target object and the distance between the target object and the intersection or the curve through the warning device further comprises:
and displaying the lane where the target object is located through warning equipment.
6. The detection and early warning method according to claim 1, further comprising:
taking a first preset time period as an interval, and storing a fog-free image with the highest definition in a plurality of fog-free images acquired within the first preset time as an analysis image;
and if the traffic accident in the first preset time period is detected, acquiring the license plate number of the vehicle in the analysis image through image recognition.
7. The detection and early warning method as claimed in claim 6, wherein if the traffic accident within the first preset time period is detected, the acquiring the license plate number of the vehicle in the analysis image by image recognition further comprises:
if the license plate number cannot be detected and obtained, the analysis image is marked and stored.
8. The detection and warning method of claim 1, wherein the second predetermined distance is greater than 20 meters.
9. A detection early warning system at a traffic intersection and a curve is characterized by comprising a camera module, a processing module and a warning module;
continuously acquiring original images at a first preset distance from a target intersection or a target curve by the camera module, wherein the camera module is arranged at an entrance of the target intersection or at the outer side of the target curve;
defogging the original image through a defogging algorithm of a countermeasure network generated based on conditions in the processing module to obtain a fog-free image; detecting a target object in the fog-free image through an image detection network in the processing module, wherein the target object comprises a vehicle and a pedestrian;
when the target object is detected, the distance between the target object and the intersection or the curve is displayed through the warning module; the warning module is arranged at a second preset distance from the target intersection or the target curve and is positioned on the inner side of the target curve.
10. The detection and early warning system according to claim 9, wherein the camera module comprises a camera provided with a holder; the warning module comprises a large screen display provided with a lamp and a loudspeaker.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110814416.7A CN113554872B (en) | 2021-07-19 | 2021-07-19 | Detection early warning method and system for traffic intersection and curve |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110814416.7A CN113554872B (en) | 2021-07-19 | 2021-07-19 | Detection early warning method and system for traffic intersection and curve |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113554872A true CN113554872A (en) | 2021-10-26 |
| CN113554872B CN113554872B (en) | 2022-09-20 |
Family
ID=78132076
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110814416.7A Active CN113554872B (en) | 2021-07-19 | 2021-07-19 | Detection early warning method and system for traffic intersection and curve |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113554872B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115661784A (en) * | 2022-10-12 | 2023-01-31 | 北京惠朗时代科技有限公司 | Intelligent traffic-oriented traffic sign image big data identification method and system |
Citations (37)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130083170A1 (en) * | 2011-10-03 | 2013-04-04 | Canon Kabushiki Kaisha | Stereoscopic image pickup apparatus |
| US20140192181A1 (en) * | 2000-03-02 | 2014-07-10 | Magna Electronics Inc. | Driver assist system for vehicle |
| EP3163501A1 (en) * | 2015-10-27 | 2017-05-03 | Fujitsu Limited | Parking space detection apparatus and method electronic equipment |
| CN108830209A (en) * | 2018-06-08 | 2018-11-16 | 西安电子科技大学 | Based on the remote sensing images method for extracting roads for generating confrontation network |
| CN109993804A (en) * | 2019-03-22 | 2019-07-09 | 上海工程技术大学 | A kind of road scene defogging method generating confrontation network based on condition |
| CN110009901A (en) * | 2019-05-24 | 2019-07-12 | 安徽超远信息技术有限公司 | Vehicle and the two-way prior-warning device of pedestrian and its application based on multiple target radar |
| CN110164128A (en) * | 2019-04-23 | 2019-08-23 | 银江股份有限公司 | A kind of City-level intelligent transportation analogue system |
| CN110310241A (en) * | 2019-06-26 | 2019-10-08 | 长安大学 | A kind of more air light value traffic image defogging methods of fusion depth areas segmentation |
| CN110322703A (en) * | 2019-07-31 | 2019-10-11 | 河北融柏信息技术服务有限公司 | One kind tests the speed bayonet type intersection safety warning system |
| CN110378849A (en) * | 2019-07-09 | 2019-10-25 | 闽江学院 | Image defogging rain removing method based on depth residual error network |
| CN110458029A (en) * | 2019-07-15 | 2019-11-15 | 佛山科学技术学院 | Vehicle checking method and device in a kind of foggy environment |
| CN110738624A (en) * | 2019-10-18 | 2020-01-31 | 电子科技大学 | area self-adaptive image defogging system and method |
| CN110781774A (en) * | 2019-10-10 | 2020-02-11 | 江苏理工学院 | Level crossing intelligent guiding system and method based on deep learning |
| US20200074239A1 (en) * | 2018-09-04 | 2020-03-05 | Seadronix Corp. | Situation awareness method and device using image segmentation |
| CN110992683A (en) * | 2019-10-29 | 2020-04-10 | 山东科技大学 | Dynamic image perception-based intersection blind area early warning method and system |
| CN111161160A (en) * | 2019-12-04 | 2020-05-15 | 新奇点企业管理集团有限公司 | Method and device for detecting obstacle in foggy weather, electronic equipment and storage medium |
| CN111415533A (en) * | 2020-04-22 | 2020-07-14 | 湖北民族大学 | Bend safety early warning monitoring method, device and system |
| EP3690809A1 (en) * | 2019-01-31 | 2020-08-05 | StradVision, Inc. | Method and device for generating safe clothing patterns for rider of bike |
| CN111507909A (en) * | 2020-03-18 | 2020-08-07 | 南方电网科学研究院有限责任公司 | Method and device for clearing fog image and storage medium |
| CN111583720A (en) * | 2020-06-08 | 2020-08-25 | 安徽达尔智能控制系统股份有限公司 | Rural highway safety early warning system based on thing networking LORA wireless technology |
| CN111738942A (en) * | 2020-06-10 | 2020-10-02 | 南京邮电大学 | Generation countermeasure network image defogging method fusing feature pyramid |
| CN111767962A (en) * | 2020-07-03 | 2020-10-13 | 中国科学院自动化研究所 | One-stage target detection method, system and device based on generation countermeasure network |
| CN111881832A (en) * | 2020-07-28 | 2020-11-03 | 苏州浪潮智能科技有限公司 | Lane target detection method, device, equipment and computer readable storage medium |
| CN111986108A (en) * | 2020-08-07 | 2020-11-24 | 西北工业大学 | Complex sea-air scene image defogging method based on generation countermeasure network |
| CN112132746A (en) * | 2020-09-17 | 2020-12-25 | 东南大学 | Small-scale pedestrian target rapid super-resolution method for intelligent roadside equipment |
| CN112150379A (en) * | 2020-09-22 | 2020-12-29 | 武汉工程大学 | Image defogging method and device for enhancing generation of countermeasure network based on perception discrimination |
| US20210019521A1 (en) * | 2018-09-04 | 2021-01-21 | Seadronix Corp. | Method and Device for Situation Awareness |
| CN112329623A (en) * | 2020-11-05 | 2021-02-05 | 山东交通学院 | Early warning method for visibility detection and visibility safety grade division in foggy days |
| CN112418027A (en) * | 2020-11-11 | 2021-02-26 | 青岛科技大学 | Remote sensing image road extraction method for improving U-Net network |
| CN112508929A (en) * | 2020-12-16 | 2021-03-16 | 奥比中光科技集团股份有限公司 | Method and device for training generation of confrontation network |
| CN112686903A (en) * | 2020-12-07 | 2021-04-20 | 嘉兴职业技术学院 | Improved high-resolution remote sensing image semantic segmentation model |
| CN112801902A (en) * | 2021-01-29 | 2021-05-14 | 福州大学 | Traffic image defogging method based on improved generation countermeasure network |
| CN112906664A (en) * | 2021-04-06 | 2021-06-04 | 昭通亮风台信息科技有限公司 | Neural network-based method and system for detecting abnormal events in tunnel |
| CN112907972A (en) * | 2021-04-06 | 2021-06-04 | 昭通亮风台信息科技有限公司 | Road vehicle flow detection method and system based on unmanned aerial vehicle and computer readable storage medium |
| CN112950458A (en) * | 2021-03-19 | 2021-06-11 | 润联软件系统(深圳)有限公司 | Image seal removing method and device based on countermeasure generation network and related equipment |
| CN113077467A (en) * | 2021-06-08 | 2021-07-06 | 深圳市华汉伟业科技有限公司 | Edge defect detection method and device for target object and storage medium |
| CN113085899A (en) * | 2021-04-28 | 2021-07-09 | 安徽江淮汽车集团股份有限公司 | Visual field blind area early warning method, device, equipment and storage medium |
-
2021
- 2021-07-19 CN CN202110814416.7A patent/CN113554872B/en active Active
Patent Citations (37)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140192181A1 (en) * | 2000-03-02 | 2014-07-10 | Magna Electronics Inc. | Driver assist system for vehicle |
| US20130083170A1 (en) * | 2011-10-03 | 2013-04-04 | Canon Kabushiki Kaisha | Stereoscopic image pickup apparatus |
| EP3163501A1 (en) * | 2015-10-27 | 2017-05-03 | Fujitsu Limited | Parking space detection apparatus and method electronic equipment |
| CN108830209A (en) * | 2018-06-08 | 2018-11-16 | 西安电子科技大学 | Based on the remote sensing images method for extracting roads for generating confrontation network |
| US20210019521A1 (en) * | 2018-09-04 | 2021-01-21 | Seadronix Corp. | Method and Device for Situation Awareness |
| US20200074239A1 (en) * | 2018-09-04 | 2020-03-05 | Seadronix Corp. | Situation awareness method and device using image segmentation |
| EP3690809A1 (en) * | 2019-01-31 | 2020-08-05 | StradVision, Inc. | Method and device for generating safe clothing patterns for rider of bike |
| CN109993804A (en) * | 2019-03-22 | 2019-07-09 | 上海工程技术大学 | A kind of road scene defogging method generating confrontation network based on condition |
| CN110164128A (en) * | 2019-04-23 | 2019-08-23 | 银江股份有限公司 | A kind of City-level intelligent transportation analogue system |
| CN110009901A (en) * | 2019-05-24 | 2019-07-12 | 安徽超远信息技术有限公司 | Vehicle and the two-way prior-warning device of pedestrian and its application based on multiple target radar |
| CN110310241A (en) * | 2019-06-26 | 2019-10-08 | 长安大学 | A kind of more air light value traffic image defogging methods of fusion depth areas segmentation |
| CN110378849A (en) * | 2019-07-09 | 2019-10-25 | 闽江学院 | Image defogging rain removing method based on depth residual error network |
| CN110458029A (en) * | 2019-07-15 | 2019-11-15 | 佛山科学技术学院 | Vehicle checking method and device in a kind of foggy environment |
| CN110322703A (en) * | 2019-07-31 | 2019-10-11 | 河北融柏信息技术服务有限公司 | One kind tests the speed bayonet type intersection safety warning system |
| CN110781774A (en) * | 2019-10-10 | 2020-02-11 | 江苏理工学院 | Level crossing intelligent guiding system and method based on deep learning |
| CN110738624A (en) * | 2019-10-18 | 2020-01-31 | 电子科技大学 | area self-adaptive image defogging system and method |
| CN110992683A (en) * | 2019-10-29 | 2020-04-10 | 山东科技大学 | Dynamic image perception-based intersection blind area early warning method and system |
| CN111161160A (en) * | 2019-12-04 | 2020-05-15 | 新奇点企业管理集团有限公司 | Method and device for detecting obstacle in foggy weather, electronic equipment and storage medium |
| CN111507909A (en) * | 2020-03-18 | 2020-08-07 | 南方电网科学研究院有限责任公司 | Method and device for clearing fog image and storage medium |
| CN111415533A (en) * | 2020-04-22 | 2020-07-14 | 湖北民族大学 | Bend safety early warning monitoring method, device and system |
| CN111583720A (en) * | 2020-06-08 | 2020-08-25 | 安徽达尔智能控制系统股份有限公司 | Rural highway safety early warning system based on thing networking LORA wireless technology |
| CN111738942A (en) * | 2020-06-10 | 2020-10-02 | 南京邮电大学 | Generation countermeasure network image defogging method fusing feature pyramid |
| CN111767962A (en) * | 2020-07-03 | 2020-10-13 | 中国科学院自动化研究所 | One-stage target detection method, system and device based on generation countermeasure network |
| CN111881832A (en) * | 2020-07-28 | 2020-11-03 | 苏州浪潮智能科技有限公司 | Lane target detection method, device, equipment and computer readable storage medium |
| CN111986108A (en) * | 2020-08-07 | 2020-11-24 | 西北工业大学 | Complex sea-air scene image defogging method based on generation countermeasure network |
| CN112132746A (en) * | 2020-09-17 | 2020-12-25 | 东南大学 | Small-scale pedestrian target rapid super-resolution method for intelligent roadside equipment |
| CN112150379A (en) * | 2020-09-22 | 2020-12-29 | 武汉工程大学 | Image defogging method and device for enhancing generation of countermeasure network based on perception discrimination |
| CN112329623A (en) * | 2020-11-05 | 2021-02-05 | 山东交通学院 | Early warning method for visibility detection and visibility safety grade division in foggy days |
| CN112418027A (en) * | 2020-11-11 | 2021-02-26 | 青岛科技大学 | Remote sensing image road extraction method for improving U-Net network |
| CN112686903A (en) * | 2020-12-07 | 2021-04-20 | 嘉兴职业技术学院 | Improved high-resolution remote sensing image semantic segmentation model |
| CN112508929A (en) * | 2020-12-16 | 2021-03-16 | 奥比中光科技集团股份有限公司 | Method and device for training generation of confrontation network |
| CN112801902A (en) * | 2021-01-29 | 2021-05-14 | 福州大学 | Traffic image defogging method based on improved generation countermeasure network |
| CN112950458A (en) * | 2021-03-19 | 2021-06-11 | 润联软件系统(深圳)有限公司 | Image seal removing method and device based on countermeasure generation network and related equipment |
| CN112906664A (en) * | 2021-04-06 | 2021-06-04 | 昭通亮风台信息科技有限公司 | Neural network-based method and system for detecting abnormal events in tunnel |
| CN112907972A (en) * | 2021-04-06 | 2021-06-04 | 昭通亮风台信息科技有限公司 | Road vehicle flow detection method and system based on unmanned aerial vehicle and computer readable storage medium |
| CN113085899A (en) * | 2021-04-28 | 2021-07-09 | 安徽江淮汽车集团股份有限公司 | Visual field blind area early warning method, device, equipment and storage medium |
| CN113077467A (en) * | 2021-06-08 | 2021-07-06 | 深圳市华汉伟业科技有限公司 | Edge defect detection method and device for target object and storage medium |
Non-Patent Citations (6)
| Title |
|---|
| ISAAC OGUNRINDE: "A Review of the Impacts of Defogging on Deep Learning-Based Object Detectors in Self-Driving Cars", 《SOUTHEASTCON 2021》 * |
| ZHONGGUO YANG: "Lane Detection by Combining Trajectory Clustering and Curve Complexity Computing in Urban Environments", 《2017 13TH INTERNATIONAL CONFERENCE ON SEMANTICS, KNOWLEDGE AND GRIDS (SKG)》 * |
| 李文杰: "应对交叉路口或弯道视野盲区的监控指示设备", 《传感器世界》 * |
| 王越: "基于金字塔注意力机制和生成对抗网络的水下图像增强", 《激光与光电子学进展》 * |
| 陈慧雅: "基于生成对抗网络的交通模糊图像复原", 《桂林电子科技大学学报》 * |
| 陈琼红: "基于AOD-Net和SSD的雾天车辆和行人检测", 《重庆理工大学学报（自然科学版）》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115661784A (en) * | 2022-10-12 | 2023-01-31 | 北京惠朗时代科技有限公司 | Intelligent traffic-oriented traffic sign image big data identification method and system |
| CN115661784B (en) * | 2022-10-12 | 2023-08-22 | 北京惠朗时代科技有限公司 | Intelligent traffic-oriented traffic sign image big data identification method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113554872B (en) | 2022-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1834583B (en) | Vehicle periphery monitoring apparatus | |
| CN110223511A (en) | A kind of automobile roadside is separated to stop intelligent monitoring method and system | |
| CN105574552A (en) | Vehicle ranging and collision early warning method based on monocular vision | |
| CN103824037B (en) | Vehicle anti-tracking alarm device | |
| DE102009048699A1 (en) | Travel's clear path detection method for motor vehicle i.e. car, involves monitoring images, each comprising set of pixels, utilizing texture-less processing scheme to analyze images, and determining clear path based on clear surface | |
| CN109389060B (en) | Vision-based vehicle surrounding collision early warning method | |
| CN107211085A (en) | Camera device and image capture method | |
| CN106295583B (en) | Method and device for reminding user of driving mode | |
| KR102397156B1 (en) | A method of providing a camera system and driver assistance functions for photographing the surrounding area of one's vehicle | |
| CN111046797A (en) | Oil pipeline warning method based on personnel and vehicle behavior analysis | |
| CN103164958B (en) | Method and system for vehicle monitoring | |
| DE112017002774T5 (en) | Image processing apparatus and image processing method, and learning device and learning method | |
| CN113554872B (en) | Detection early warning method and system for traffic intersection and curve | |
| CN110115024A (en) | Image processing apparatus, image processing method and program | |
| JP2013182550A (en) | Vehicle detection device, vehicle detection method and vehicle detection program | |
| DE112015005633T5 (en) | All-round visibility system and vehicle that includes a all-round visibility system | |
| CN110929606A (en) | Vehicle blind area pedestrian monitoring method and device | |
| CN115134537A (en) | Image processing method and device and vehicle | |
| CN111241918B (en) | Vehicle tracking prevention method and system based on face recognition | |
| CN106991415A (en) | Image processing method and device for vehicle-mounted fisheye camera | |
| CN111126237A (en) | Safe vehicle distance detection method based on machine vision | |
| CN207115438U (en) | Image processing apparatus for vehicle-mounted fisheye camera | |
| KR101992652B1 (en) | Method and apparatus for detecting vehicle through detection of license plate | |
| JP2019192201A (en) | Learning object image extraction device and method for autonomous driving | |
| CN115240152A (en) | Road property and road right vehicle-mounted identification system based on video deep learning technology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |