CN109147347A - Parallel parking the night detection of vehicle - Google Patents

Abstract

The present invention proposes a kind of night detection system for paralleling parking vehicle.It is illuminated and is detected using the car light of passing cars and parked cars.The system contains one and parks cars sensor and a moving vehicle sensor.The sensor that parks cars monitors a parking area (containing multiple parking stalls), and moving vehicle sensor perceives the moving vehicle near the parking area.After moving vehicle sensor perceives an at least passing cars, the sensor that parks cars extracts the echo area of at least a headlight or at least partly Guard Assy.

Description

Parallel parking the night detection of vehicle

Technical field

The present invention relates to electronic fields, more precisely, being related to the monitoring to parking stall.

Background technique

Drivers often feel put about because can not find parking stall.The process wastes time on parking stall is found, combustion is increased Oil consumption has negative effect to environment.In order to energy saving, and environmental quality is improved, there is an urgent need to develop a kind of parking stall prisons Examining system, the occupancy situation on parking stall can be monitored in real time in it.After obtaining parking bit occupancy data, driver can promptly exist Nearby find parking stall in destination.

Parking stall law enforcement is an important component of city management.Existing parking stall enforcement system is to pass through parking Law enfrocement official carries out permanent patrol to parking street and parking lot to realize.This patrol mode needs to consume a large amount of manpowers Material resources.In order to avoid this waste, it is desirable to monitor in each parking stall of systematic survey to be stopped using above-mentioned parking stall Down time.

Parking stall monitoring and parking stall law enforcement are required to detect and park cars.Park cars detection preferably on daytime and Night can implement.Such as shopping centre is more focused on daytime observation, residential quarter is more focused on night detection.Due to previous skill Art is illuminated using available light, they can only monitor in the daytime parking stall state.At night, due to street lamp cannot provide can The illumination (having tree or other barrier obstructions) leaned on, conventional art are difficult to monitor parking stall state at night.

Summary of the invention

The main object of the present invention is energy saving, and improves environmental quality.

It is another object of the present invention to the detections at night to park cars.

It is another object of the present invention to realize that parking at night position monitors automatically.

It is another object of the present invention to realize that parking at night position is enforced the law automatically.

According to these and other purpose, the present invention proposes a kind of night detection system to park cars.It utilizes and passes by The car light of vehicle parks cars to illuminate and detect.The system contains one and parks cars sensor and a moving vehicle sensor. The sensor that parks cars monitors a parking area (containing multiple parking stalls), and the perception of moving vehicle sensor is attached in the parking area Close moving vehicle,.After moving vehicle sensor perceives an at least passing cars, the sensor that parks cars shoots and handles this The photo of parking area.

The car light of passing cars is limited in scope (about 20 meters), it can only illuminate a small number of parking vehicles (general 3).Due to The time that passing cars pass by is very short (within several seconds), and the sensor that parks cars needs take a picture at least every 2 seconds.This Filming frequency is more much greater (every 5-10 second beats one is opened in the daytime) than filming frequency in the daytime.When the processor for the sensor that parks cars When very powerful, these photos can be handled in real time；Otherwise first photo can be stored, is reprocessed after passing cars leave These photos.

Due to the vehicle light illumination to park cars by passing cars, rather than available light, the processing of nighttime image and daytime are not Together.Firstly, the image ROI(region of interest at night, the region of image procossing is carried out to the photo on the parking stall) From it is different in the daytime, it has different shape and position.Secondly, night extract vehicle characteristics from it is different in the daytime.What night extracted Vehicle characteristics are mainly reflecting region (pixel intensity is stronger), and the vehicle characteristics extracted in the daytime are mainly edge (pixel intensity Change stronger).For sequence parked vehicle, (vehicle body of all vehicles is parked along a line, and the photo of shooting is these vehicles Side view), night extract feature be mainly taillight echo area, wheel echo area and vehicle body echo area etc..For stopping in parallel The vehicle (vehicle body parallel of all vehicles, the photo of shooting are the headstock or the tailstock of these vehicles) put, the feature that night extracts Mainly head lamp/taillight echo area, front/rear buffer arm echo area etc..

Detailed description of the invention

Fig. 1 is the top view of a street: vehicle is parked along curbside, and a moving vehicle is just passing by these and parking cars.

Fig. 2 is a kind of block diagram of night detecting devices that parks cars.

Fig. 3 is a kind of block diagram of sensor that parks cars.

Fig. 4 A- Fig. 4 C shows multi-motion vehicle sensors and moving vehicle cognitive method.

Fig. 5 A- Fig. 5 B is the flow chart of two kinds of night probe algorithms that park cars.

Fig. 6 indicates that the sequence extracted in the daytime parks cars feature (conventional art).

The sequence that Fig. 7 indicates that night extracts parks cars feature.

What Fig. 8 display night extracted parallels parking vehicle characteristics.

It is noted that these attached drawings are only synoptic diagrams, their nots to scale (NTS) are drawn.For the sake of obvious and is convenient, in figure Portion size and structure may zoom in or out.In different embodiments, identical symbol typicallys represent corresponding or similar Structure.

Specific embodiment

Fig. 1 is an avenue 20, it is along x axis, and there are two curbside 20a, 20b.Along curbside, 20a has multiple parking stalls 10a-10f；It is deployed with a parking stall monitoring device 30a in opposite curbside 20b, equipment 30a monitoring parking area 35(includes Parking stall 10a-10f).It is, in general, that equipment 30a is deployed on electric pole or lamppost, these electric poles or lamppost are also Equipment 30a provides power supply.It parks cars for ease of detection, the height of equipment 30a deployment, which is preferably higher than, to park cars.

In the parking area 35 of monitoring, 4 parking stalls 10a, 10c, 10d and 10f have been accounted for, they parked vehicle 40a, 40c, 40d and 40f；Other 2 parking stalls 10b, 10e are not accounted for.At daytime (available light illumination under), parking stall 10a-10f's State is easily monitored by parking stall monitoring device 30a.At night, since parking stall 10a-10f does not have enough illuminations, one The car light 60 of moving vehicle 50 is used to illumination and parks cars, to determine the state of parking stall 10a-10f.

Fig. 2 is a kind of block diagram of parking stall monitoring device 30.It is illuminated at night using the car light 60 of moving vehicle 50 It parks cars.The equipment 30 parks cars sensor 80 and a moving vehicle sensor 70 containing one.Park cars sensor 80 A parking area 35(is monitored referring to Fig. 3).Moving vehicle sensor 70 perceives at least passing cars 50(ginseng, and A- schemes referring to fig. 4 4C).Passing cars 50 refer to the moving vehicle from parking area 35 less than a pre-determined distance.

Fig. 3 is a kind of block diagram of sensor 80 that parks cars.It contains a camera 82, a processor 84 and a memory 86.Camera 82 shoots parking area 35.Camera 82 can have it is several, and face different directions.Processor 84 processing by The photo of the shooting of camera 82 simultaneously determines dead ship condition.Memory 86 stores the photo at least partly shot by camera 82, with And the probe algorithm 87 that parks cars.

Fig. 4 A- Fig. 4 C shows multi-motion vehicle sensors 70 and moving vehicle cognitive method.In Figure 4 A, moving vehicle Sensor 70 can be an acoustic sensor, optical sensor or electromagnetic sensor.It is attached that acoustic sensor monitors parking area 35 Sound variation caused by nearly moving vehicle 50；Optical sensor (camera in the sensor 80 that such as parks cars) perceives parking area The variation of light intensity caused by moving vehicle 50 (B referring to fig. 4) near 35；Electromagnetic sensor detects parking area using modes such as antennas Electromagnetic field caused by moving vehicle 50 changes near domain 35.

Fig. 4 B uses the sensor 80 that parks cars in Fig. 3 as moving vehicle sensor 70.Park cars sensor 80 Memory 86 a moving vehicle probe algorithm 89 is also stored.The light that the algorithm 89 monitors the neighbouring road surface that parks cars is strong Degree, once the variation of light intensity is more than a threshold value, then it represents that and moving vehicle enters parking area.

In figure 4 c, the sensor 80a that parks cars that the moving vehicle sensor 70b of neighbouring block 22b is this block 22a It is provided with prenoticing for moving vehicle arriving.Wireless signal 98(such as WiFi, bluetooth can be used in moving vehicle sensor 70b Deng) notice passed into the sensor 80a that parks cars.It is noted that moving vehicle sensor 70b and the sensor that parks cars 80a is the component of parking stall monitoring device in each block respectively.In the case where prenoticing, park cars sensing Device 80a can more effective monitor parked vehicle.

Fig. 5 A- Fig. 5 B is the flow chart of two kinds of night probe algorithms that park cars.In fig. 5, the sensor that parks cars is clapped The photo taken the photograph is processed when passing cars 50 pass by；In figure 5B, the photo of the sensor that parks cars shooting is in passing cars 50 leave it is rear processed.

As shown in Figure 5A, the first night probe algorithm that parks cars includes the following steps.Moving vehicle sensor 70 is felt Know a moving vehicle 50(step 110).If moving vehicle 50 is located at (i.e. the passing cars) (step nearby of parking area 35 120) photo (step 130) of the sensor 80 that, parks cars shooting parking area 35.It is all to be stopped by passing cars vehicle light illumination Parking stall image (step 140) processed.Step 130,140 are repeated, until passing cars 50 leave parking area 35(step 150).Then next passing cars (step 160) is waited.

The car light of passing cars is limited in scope (about 20 meters), it can only illuminate a small number of parking vehicles (general 3).Due to The time that passing cars pass by is very short (within several seconds), and the sensor that parks cars needs take a picture at least every 2 seconds.This Filming frequency is more much greater (every 5-10 second beats one is opened in the daytime) than filming frequency in the daytime.When the processing for the sensor 80 that parks cars When device 84 is very powerful, these photos can be handled in real time；Otherwise first photo can be stored, leaves and stops in passing cars 50 These photos are reprocessed behind vehicle region 35.This is further described in figure 5B.When passing cars 50 are near parking area 35 (step 120), the sensor 80 that parks cars only shoots photo (step 130), and memory 86(step 145) is recorded in they In, but do not handle these photos.(the step 150) after passing cars 50 leave parking area 35, processor 84 reprocess this A little photos and the state (step 155) for determining parking stall.

Since the car light 60 to park cars by passing cars illuminates, rather than available light, the processing of nighttime image and daytime It is different.Fig. 6 and Fig. 7 compares the difference with nighttime image processing in the daytime, the difference of especially ROI and vehicle characteristics.

Fig. 6 shows ROI 200a in the daytime, 200c along curbside 20a, vehicle 40a, 40c for being parked in parking stall 10a, 10c. Since they are parked along curbside 20a, these vehicles 40a, 40c are that sequence parks cars.Each ROI(such as 200a) it begins approximately even with and stops Parking stall (such as 10a) exposure sideline (as " ab "), scanning up to is more than at least side window of (such as 40a) of parking cars.ROI In vehicle characteristics be vehicle edge feature.For sequence parked vehicle, edge feature includes vehicle body lower edge and side Window lower edge etc. (referring to United States Patent (USP) 8,923,565).

Fig. 7 display order parks cars the night ROI of 40a, 40c.Each vehicle (such as 40a) there are two ROI(such as 210a, 220a).First ROI 210a covers the wheel and at least part vehicle body for the 40a that at least parks cars.2nd ROI 220a Cover at least taillight for the 40a that parks cars.Night extract vehicle characteristics from it is different in the daytime: night feature is mainly echo area Domain (pixel intensity is stronger), and feature is mainly edge in the daytime (pixel intensity variation is stronger).For sequence parked vehicle, night Between the vehicle characteristics that extract be mainly wheel echo area 320, taillight echo area 330 and vehicle body echo area 340 etc..Here, feature Echo area can be more than the pixel of a threshold value obtain by searching brightness in the roi.

Fig. 8 shows the night ROI for paralleling parking vehicle 40d, 40f.Vehicle 40d, 40f be parked in respectively parking stall 10d, 10f, and parking stall 10e is unoccupied.Each vehicle (such as 40d) has a ROI(such as 230d).ROI 230d covers vehicle 40d Two headlights and at least partly before buffer arm.Parallel parking vehicle 40d night feature include its headlight reflection area 350d, 360d and preceding buffer arm echo area 370d etc..Similarly, the night feature for paralleling parking vehicle 40f includes its headlight reflection area 350f, 360f and preceding buffer arm echo area 370f etc..

It should be appreciated that under the premise of not far from the spirit and scope of the present invention, it can be to form and details of the invention It is modified, this does not interfere them using spirit of the invention.Therefore, in addition to the spirit according to appended claims, The present invention should not be any way limited.

Claims (10)

1. a kind of night detecting devices for paralleling parking vehicle, it is characterised in that contain:

One parks cars sensor, which monitors a parking area；

One moving vehicle sensor, the moving vehicle sensor perceive the moving vehicle near the parking area；

After the moving vehicle sensor perceives an at least passing cars, which extracts an at least headlight or extremely The echo area of small part Guard Assy.

2. equipment according to claim 1, it is further characterized in that: the passing cars are at a distance from the parking area less than one Pre-determined distance.

3. equipment according to claim 1, it is further characterized in that: the moving vehicle sensor contain an acoustic sensor, At least one of optical sensor or electromagnetic sensor.

4. equipment according to claim 1, it is further characterized in that: the sensor that parks cars contain an at least camera, One processor and a memory.

5. equipment according to claim 1, it is further characterized in that: park cars sensor and the moving vehicle sensor It is co-located.

6. equipment according to claim 1, it is further characterized in that: park cars sensor and the moving vehicle sensor Positioned at different location.

7. equipment according to claim 1, it is further characterized in that: park cars sensor and the moving vehicle sensor Pass through a wireless communication.

8. equipment according to claim 1, it is further characterized in that: the sensor that parks cars is big in the frequency of shooting at night In the frequency shot in the daytime.

9. equipment according to claim 1, it is further characterized in that: this parks cars sensor when the passing cars pass by Handle the photo.

10. equipment according to claim 1, it is further characterized in that: the sensor that parks cars is passed by the passing cars Post-process the photo.