Vehicle detector for detecting microwave and video fusion
Technical Field
The utility model relates to a vehicle detector that microwave and video integration detected.
Background
There are three main types of vehicle detectors on the current road: geomagnetic-type vehicle detectors, microwave-type vehicle detectors, and video-type vehicle detectors.
The geomagnetic vehicle detector needs to cut a road for paving and burying construction, is difficult to construct and high in construction cost, and once the road needs to be constructed again (maintained or changed), the geomagnetic vehicle detector is damaged. Metal substances near the road surface have a significant influence on the geomagnetic vehicle detector. A geomagnetic vehicle detector can only detect the traffic flow condition of a certain section on a single lane, and the provided road information is limited.
The microwave vehicle detector emits microwave beams to the road surface of the detection area, the vehicle generates reflection signals when entering the beam irradiation range, and the microwave vehicle detector receives the reflection signals of the vehicle for processing to obtain the position and speed information of the vehicle. The microwave type vehicle detector can detect a section of a road, and is less affected by weather conditions (day and night performance is not different, and influence of rain, snow, wind and frost is less). However, the microwave vehicle detector cannot provide visual information of a road site, cannot detect a stopped vehicle, often detects a false target behind a large vehicle, cannot effectively distinguish between the detection of vehicles at close intervals, and cannot effectively distinguish between the large vehicle, a small vehicle, a non-motor vehicle and a pedestrian.
The video type vehicle detector obtains the position information of the vehicle on the picture by carrying out image processing and analysis on the road surface video of the monitored area, and can roughly estimate the speed of the vehicle according to the change of the picture position. The video-type vehicle detector is limited in detection and identification capability and can only perform better identification on the central region of the picture. The video vehicle detector has the advantages that the video vehicle detector can provide the image condition of the road surface, and has excellent detection performance on the large vehicles and the dense small vehicles in the central area of the image under the ideal brightness condition; the identification effect on different vehicle types and pedestrians is good. The video type vehicle detector has disadvantages in that it is greatly affected by a bright condition and a weather condition (performance is deteriorated at night, and it is basically impossible to work in a rainy and foggy weather).
In summary, the three vehicle detectors commonly used at present have advantages and disadvantages, but do not provide a good enough detection effect.
SUMMERY OF THE UTILITY MODEL
The utility model aims to overcome above-mentioned prior art not enough, if an equipment, can adopt microwave and video to detect simultaneously, and the data of these two kinds of detection methods can be unified each other, this equipment will provide the detection effect of fairly ideal, can guarantee enough big detection area by microwave detection, and not influenced by the weather condition, can provide on-the-spot influence information by video detection again, accomplishes static target, the cart detects and distinguishes various motorcycle type categories.
Target information detected by the microwave vehicle detector is based on the real world, and the coordinates and the speed of the target are based on the coordinates (a world coordinate system) and the speed of the real world; the target information detected by the video vehicle detector is based on the picture, and the coordinates of the target are pixel coordinates of the picture. How to unify the real world-based target information of the microwave vehicle inspection device and the picture pixel-based information of the video vehicle inspection device in the same vehicle detector and install and use the same on an actual road is a great problem. The utility model provides a vehicle detector that microwave and video integration detected, it is big to have detection area, can provide accurate target location, speed information and on-the-spot image information, can detect the effect of the target that stops.
The utility model discloses a realize like this, a vehicle detector that microwave and video fuse detected, including the shell main part, panel fixed frame and hou gai are installed respectively to shell main part both sides, be provided with microwave protection panel between panel fixed frame and the shell main part, be provided with camera, microwave detection module, power strip and processing module in the shell main part, be equipped with the round hole in the microwave protection panel that the camera corresponds, the power strip respectively with camera, microwave detection module and processing module electric connection, processing module respectively with camera and microwave detection module data connection, the shell main part outside is provided with the installing support.
Preferably, the shell main body is an aluminum alloy shell main body, a partition plate is arranged in the middle of the shell main body, the camera and the microwave detection module are arranged at the front end of the partition plate, and the power panel and the processing module are arranged at the rear end of the partition plate.
Preferably, a plurality of heat dissipation holes are formed in the outer side of the middle of the shell body.
Preferably, the mounting bracket is hinged to the housing body.
Preferably, the method for mounting the vehicle detector includes the steps of:
1) preliminarily mounting the vehicle detector at a specified mounting position through a mounting bracket, preliminarily adjusting the azimuth angle and the pitch angle of the vehicle detector according to a video image output by a camera, aligning the vehicle detector to the area to be detected, and judging that the center of the video image is overlapped with the center of the area to be detected according to the judgment basis;
2) then starting a microwave detection module, drawing target data detected by the microwave detection module into a track heat map, adjusting an azimuth angle of a vehicle detector according to the shape of the track heat map to enable the track heat map to appear vertically, enabling a vehicle sensor to be opposite to the direction of a lane at the moment, adjusting the azimuth angle, adjusting a pitch angle, enabling the microwave detector to detect target tracks at the far end and the near end of a detection area, and adjusting the pitch angle;
3) after the azimuth angle and the pitch angle are adjusted, fixing the vehicle detector;
4) after the vehicle detector is fixed, drawing a target track heat map again, drawing a road map of a detection area according to the shape of the track heat map, referring to the detection area map, matching the positions of a target detected by a microwave detection module and the target in a camera picture, acquiring pixel coordinates of the target in the picture, performing 12 groups of target matching, acquiring 12 groups of pixel coordinates of the target, acquiring world coordinates of the corresponding target by combining detection of the microwave detection module, forming a coordinate conversion equation set, solving the equation set, obtaining homography matrix data of the camera under the current installation condition, and transmitting the homography matrix data to the inside of a processing module of the vehicle detector for a tracking algorithm;
5) the microwave detector is installed and debugged, and can be normally used.
Preferably, the homography matrix data of the camera in the step 4) is used for realizing the interconversion between target data based on a world coordinate system and obtained by the built-in microwave detection module of the vehicle detector and target data based on picture pixel coordinates and obtained by the camera.
Preferably, the method for acquiring the homography matrix data of the camera in the step 4) includes: the conversion relation calculation formula of the target pixel coordinate of the camera and the world coordinate of the target in the real world is as follows:
wherein (Xp, Yp, Zp) is the pixel coordinate in the camera picture, (Xw, Yw, Zw) is the coordinate of the target in the world coordinate system, and the H matrix calculation relation is:
when the vehicle detector is installed at the designated position, determining a homography matrix of the camera, detecting target data by collecting a microwave detection module, drawing the target data into a heat map, showing the lane condition of a detection area by the heat map, drawing a map of the detection area according to the lane condition of the heat map, obtaining a detection area map, comparing a target detected by the microwave detection module in the detection area map with a target in a camera picture, completing the matching of the target detected by the microwave detection module and the target in the camera picture, extracting pixel coordinates (Xp, Yp, Zp) of the target after the target in the picture is matched with the target detected by the microwave detection module, combining the coordinate conversion relation and the coordinates (Xw, Yw, Zw) of the given target detected by the microwave detection module based on a world coordinate system, an equation with an H matrix as an unknown number is obtained, the H matrix is composed of 12 elements, 12 pixel coordinates and world coordinates of successfully paired targets are selected, an equation set can be composed of 12 equations, and the H matrix can be obtained through solving.
Preferably, the tracking algorithm in step 4) is: in each detection period, after target data of a microwave detection module and target data detected by a camera are obtained, checking the validity of the data to remove invalid data, then matching the two types of target data obtained by detection with the tracked target according to the position, the speed and the size of the vehicle model, and if the two types of target data are matched, indicating that the detected data belong to the tracked target, and updating the tracked target by using newly detected data; after the tracked targets are matched, effective data which are still not matched are established, and the tracked targets are established according to the effective data; and acquiring final target data by adopting a method of weighting the target data acquired by detecting the target data by the microwave detection module and the target data by the camera and the reliability degree, and updating the target tracking state by using the final target data.
Preferably, the reliability level includes the following conditions: the target is detected by the target data of the microwave detection module and the target data of the camera, and the reliability is high; the target data of the microwave detection module and the target data of the camera detection detect the target, and the reliability degree is different; one of the target data of the microwave detection module and the target data of the camera detects the target; the target is not detected by the target data of the microwave detection module and the target data of the camera.
To sum up, owing to adopted above-mentioned technical scheme, the beneficial effects of the utility model are that: by adopting the microwave and video fused vehicle detection equipment, after being installed, the advantages of the microwave type vehicle detector and the advantages of the video type vehicle detector can be combined, and the respective disadvantages of the microwave type vehicle detector and the video type vehicle detector are eliminated; thereby having the following advantages: the detection area is large and is hardly influenced by weather conditions, and the defects that the detection area of single video detection is insufficient and is greatly influenced by the weather conditions are overcome. Accurate target position and speed information can be provided, and the defect that the accurate target position and speed cannot be provided by single video detection is overcome. The system can provide on-site image information, can well detect two vehicles with big vehicles and small space and vehicles of various different vehicle types, and overcomes the defects of poor detection effect of the independent microwave detection big vehicle and the vehicles with small space and poor vehicle type distinguishing effect. The stopped target can be detected, and the defect that a static target cannot be detected by microwave detection alone is eliminated. These performances can be obtained simultaneously, which were not available in the prior art with separate microwave-type vehicle detectors and video-type vehicle detectors.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is an explosion structure diagram of a vehicle detector for detecting microwave and video fusion according to an embodiment of the present invention.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
As shown in FIG. 1, a vehicle detector for microwave and video fusion detection is mounted on a road-side vertical rod, an L-shaped frame and a portal frame. The vehicle detector detects motor vehicles, non-motor vehicles and pedestrians in a detection area by adopting a method of fusing microwave detection and video detection. The microwave detection device comprises a shell main body 5, a panel fixing frame 1 and a rear cover 8 are respectively installed on two sides of the shell main body 5, a microwave protection panel 2 is arranged between the panel fixing frame 1 and the shell main body 5, a camera 3, a microwave detection module 4 (the principle of the microwave detection module is that a device for detecting a target is realized by measuring microwave basic parameters such as amplitude attenuation, phase shift or frequency change according to the change of physical properties such as microwave reflection, transmission, diffraction coherent radiation, cavity perturbation and the like and the relative change of a dielectric constant and a loss tangent angle of the detected target), a power panel 6 and a processing module 7 are arranged in the shell main body 5, the camera 3 and the microwave detection module 4 are integrally assembled on the same plastic shell, wherein the camera 3 acquires image information, and the processing module 7 operates an artificial intelligent target identification algorithm to acquire image target information, the processing module 7 (in particular using an existing data processor) also assumes the full algorithmic and software functionality of the entire vehicle detector. The microwave protection panel 2 corresponding to the camera 3 is internally provided with a circular hole 10, the microwave protection panel 2 is formed by injection molding of a Polycarbonate (PC) material, and the microwave protection panel is excellent in microwave penetration characteristic, excellent in ultraviolet resistance and weather resistance and suitable for application scenes of outdoor use of a vehicle detector. The circular hole 10 of the microwave protection panel 2 is embedded with an acrylic sheet with excellent light transmittance as a window of the camera 3. The power panel 6 is respectively electrically connected with the camera 3, the microwave detection module 4 and the processing module 7, the processing module 7 is respectively connected with the camera 3 and the microwave detection module 4 in a data mode, and the mounting support 9 is arranged on the outer side of the shell main body 5. The shell main body 5 is an aluminum alloy shell main body, and high-heating parts in the camera 3, the microwave detection module 4 and the processing module 7 are directly or indirectly (by taking a good heat dissipation conductor as an intermediate medium) tightly attached to the shell main body 5, so that heat is conducted to the shell main body 5 to be dissipated. The inside baffle 11 that is provided with in 5 middle parts of shell main part separates into preceding chamber and back chamber with shell main part 5, camera 3 and microwave detection module 4 set up at 11 front ends of baffle, power strip 6 and processing module 7 set up at 11 rear ends of baffle. The outer side of the middle of the shell main body 5 is provided with a plurality of heat dissipation holes 12, the heat dissipation holes 12 are a plurality of vertical through holes and horizontal through holes, the overall heat dissipation performance is improved, and the artificial intelligence target recognition algorithm with large operation power consumption of the video detection module is facilitated. The outer edges of the two ends of the shell main body 5 are provided with waterproof rubber strips, and the waterproof rubber strips are attached and compressed when the panel fixing frame 1, the microwave protection panel 2, the rear cover 8 and the shell main body 5 are integrally installed, so that the waterproof performance of outdoor use is ensured. The mounting bracket 9 is hinged to the housing body 5 so that the angle of the housing body 5 can be adjusted.
After the vehicle detector comprises the microwave detection function and the video detection function, further, data obtained by the two detection modules need to be mutually converted and used by an algorithm together. Converting target data based on world coordinates detected by a microwave detection module into a camera picture to obtain corresponding pixel coordinates, wherein a coordinate conversion equation is required to be used:
converting the target world coordinates (Xw, Yw, Zw) obtained by the microwave detection module into pixel coordinates in a picture, wherein an H matrix is required to be obtained:
i.e. the homography of the camera.
Aiming at the characteristic that the vehicle detector is internally provided with the microwave detection module and the camera at the same time, the target obtained by the microwave detection module can be matched with the corresponding target in the picture by a manual method, so that pixel coordinates (Xp, Yp, Zp) are obtained. With the world coordinates (Xw, Yw, Zw) and pixel coordinates (Xp, Yp, Zp) of the same object, a transformation equation can be listed. The H matrix is composed of 12 elements, 12 pixel coordinates and world coordinates of successfully paired targets are selected, an equation set can be composed of 12 equations, and the H matrix can be obtained through solving.
When the microwave detection is finished to obtain the matching of the target and the corresponding target in the picture, a detection area map needs to be drawn firstly, and the matching of the microwave detection target and the corresponding target in the picture can be quickly and effectively finished according to the comparison reference of the lane condition in the detection area map and the lane condition in the picture. The target data are drawn into a heat map by collecting more microwave detection target data, the heat map shows the lane condition of the detection area, and a map of the detection area can be drawn according to the lane condition of the heat map.
After the on-site map is obtained, the radar detection target and the position of the target in the picture can be conveniently matched, so that the pixel coordinate is obtained.
After the vehicle detector is mounted and the homography matrix of the camera at the mounting position is obtained, the world coordinate data and the pixel coordinate data of the object can be mutually converted according to the conversion relation. Ideally, after the world coordinates of the same target are converted into picture pixel coordinates, the picture pixel coordinates should coincide with pixel coordinates obtained by video detection, and similarly, after the pixel coordinates are converted into the world coordinates, the picture pixel coordinates should also coincide with world coordinates obtained by microwave detection. That is, the vehicle detector may obtain two kinds of data by microwave detection and video detection, and the two kinds of data may be converted and then processed without difference. The processing method of the two data comprises the following steps: in each detection period, after the microwave detection target data and the video detection target data are obtained, the data validity is checked, and invalid data, such as target data on non-roads given in video detection, are removed. And then, matching the two types of detected target data with the targets which are tracked, wherein the matching basis can be conditions such as position, speed, vehicle type and the like. If the matching is positive, the detected data belongs to the tracked target, and the tracked target can be updated by the newly detected data. After all the tracked targets are matched, the effective data which are not matched are still matched, and the tracked targets can be newly built according to the data. For the same target, several matching situations occur: the target is detected by both the microwave and the video, and the reliability is high; the target is detected by both the microwave and the video, but the credibility has high and low differences; only one of the microwave and the video detects the target; neither the microwave nor the video detects the target. At the moment, the target data obtained by the two detection methods and the final target data are obtained by a reliability weighting method, and the target tracking state is updated by the final target data.
And finally, installing and debugging the vehicle detector. The vehicle detector is preliminarily mounted to a specified mounting position, and the fixing screw is not tightened for the moment. According to the video image output by the built-in camera, the azimuth angle and the pitch angle of the vehicle detector are preliminarily adjusted, so that the vehicle detector is aligned to the area to be detected, and the center of the video image is approximately overlapped with the center of the area to be detected according to the judgment basis. And then, starting the microwave detection module, drawing target data detected by the microwave detection module into a heat map, finely adjusting the azimuth angle of the vehicle detector according to the condition of the heat map to enable the track heat map to appear vertically, showing that the vehicle sensor is opposite to the lane direction at the moment, adjusting the azimuth angle, and finely adjusting the pitch angle to enable the microwave detector to detect the target tracks at the farthest end and the nearest end of the detection area, and adjusting the pitch angle. And after the adjustment of the azimuth angle and the pitch angle is finished, completely fixing the vehicle detector. After the fixation, the target track heat map at that time is drawn again, and a road map of the detection area is drawn based on the shape of the track heat map. And then, referring to the detection area map, matching the microwave detection target with the position of the target in the camera picture, and acquiring the pixel coordinate of the target in the picture. Matching 12 groups of targets, acquiring pixel coordinates of the 12 groups of targets, obtaining world coordinates of corresponding targets by combining microwave detection, forming a coordinate conversion equation set, solving the equation set, solving a homography matrix of the built-in camera under the current installation condition, and updating the homography matrix into the vehicle detector for algorithm use. Therefore, the microwave detector is installed and debugged and can be normally used.
To sum up, owing to adopted above-mentioned technical scheme, the beneficial effects of the utility model are that: by adopting the microwave and video fused vehicle detection equipment, after being installed, the advantages of the microwave type vehicle detector and the advantages of the video type vehicle detector can be combined, and the respective disadvantages of the microwave type vehicle detector and the video type vehicle detector are eliminated; thereby having the following advantages: the detection area is large and is hardly influenced by weather conditions, and the defects that the detection area of single video detection is insufficient and is greatly influenced by the weather conditions are overcome. Accurate target position and speed information can be provided, and the defect that the accurate target position and speed cannot be provided by single video detection is overcome. The system can provide on-site image information, can well detect two vehicles with big vehicles and small space and vehicles of various different vehicle types, and overcomes the defects of poor detection effect of the independent microwave detection big vehicle and the vehicles with small space and poor vehicle type distinguishing effect. The stopped target can be detected, and the defect that a static target cannot be detected by microwave detection alone is eliminated. These performances can be obtained simultaneously, which were not available in the prior art with separate microwave-type vehicle detectors and video-type vehicle detectors.
The foregoing is a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, a plurality of improvements and decorations can be made without departing from the principle of the present invention, and these improvements and decorations are also considered as the protection scope of the present invention.