CN112016517A - Parking space identification method and device based on machine vision - Google Patents
Parking space identification method and device based on machine vision Download PDFInfo
- Publication number
- CN112016517A CN112016517A CN202010958401.3A CN202010958401A CN112016517A CN 112016517 A CN112016517 A CN 112016517A CN 202010958401 A CN202010958401 A CN 202010958401A CN 112016517 A CN112016517 A CN 112016517A
- Authority
- CN
- China
- Prior art keywords
- parking space
- array
- state
- image
- parking
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention provides a parking space identification method and device based on machine vision, and relates to the technical field of monitoring management systems. The parking space identification method based on the machine vision comprises the following steps: designing sampling points on the parking places in the parking lot; processing the image; acquiring initial information values of the plurality of sampling points, and acquiring judgment of the background array on the state of the background array by the background array to acquire a sampling point state array; judging the state of each parking space according to the state array; updating the background array; and updating the distribution map of the vacant parking spaces. The parking space recognition device based on the machine vision comprises a memory and a processor; the memory is used for storing computer executable instructions; the processor is used for executing the instructions of the computer executable method. The invention ensures that the vehicle can see the distribution of the vacant parking spaces before driving into the parking lot, and the vehicle can directly drive into the vacant parking spaces without causing vehicle congestion.
Description
Technical Field
The invention relates to the technical field of monitoring management systems, in particular to a parking space identification method and device based on machine vision.
Background
With the rapid development of economy, the quantity of motor vehicles kept by residents in China is continuously increased, and cities face more and more serious traffic problems. Vehicle congestion is the main embodiment of the current urban traffic problem, wastes a large amount of energy, pollutes the urban environment and seriously affects the traveling quality of people.
In the traditional mode, only statistics is carried out on whether vacant parking spaces and the number of remaining parking spaces exist, and the positions of the specific remaining parking spaces are not displayed. Since no special parking space induction system exists on public roads, many vehicles continuously roam to find parking spaces, so that the urban traffic flow is overlarge and congestion is caused.
Disclosure of Invention
The invention provides a parking space identification method and device based on machine vision, and aims to solve the technical problems of difficulty in parking and congestion caused by parking.
The technical scheme for realizing the invention is as follows:
the invention provides a parking space identification method based on machine vision, which comprises the following steps:
designing sampling points on the parking places in the parking lot;
processing the image;
acquiring initial information values of the plurality of sampling points, and acquiring judgment of the background array on the state of the background array by the background array to acquire a sampling point state array;
judging the state of each parking space according to the state array;
updating the background array;
and updating the distribution map of the vacant parking spaces.
Optionally, after the determining the state of the background array, the method further includes: shape filtering the array of states.
Optionally, the performing shape filtering on the state array includes:
converting the image to a frequency domain;
different digital filters are designed to filter the image.
Optionally, designing a plurality of sampling points in the area image of the parking space includes: the positions and the number of the sampling points and the areas of the sampling points are designed.
Optionally, the determining the state of each parking space according to the state array includes: the image is processed using a method of image histograms.
Optionally, the information values comprise pixel values.
Optionally, the updating the background array includes: and repeating the steps of obtaining the real-time information values of the plurality of sampling points, judging the real-time information values to obtain a state array, and judging the state of each parking space according to the state array.
The invention provides a parking space recognition device based on machine vision, which comprises a memory and a processor, wherein the memory is used for storing parking space data; the memory is used for storing computer executable instructions; the processor is used for executing the instructions of the computer executable method.
Optionally, the parking space management system further comprises a display, and the display receives the result of the processor and displays the state of each parking space.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a plan recognition device and method based on machine vision, which can detect parking spaces and detect the distribution of the parking spaces in a vacant area in real time, so that a vehicle can see the distribution of vacant parking spaces before driving into a parking lot, and the vacant parking spaces are directly driven into the vacant parking spaces without causing vehicle congestion, energy waste and environmental pollution.
Drawings
Fig. 1 is a flowchart of a parking space identification method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of distribution of sampling points when a vehicle is parked in a parking space according to an embodiment of the present invention;
fig. 3 is a schematic view of distribution of sampling points in a parking space provided by the embodiment of the invention when the parking space is empty.
Reference numerals: 1-sampling point; 2-parking space.
Detailed Description
The present invention is described in detail with reference to the embodiments shown in the drawings, but it should be understood that these embodiments are not intended to limit the present invention, and those skilled in the art should understand that functional, methodological, or structural equivalents or substitutions made by these embodiments are within the scope of the present invention.
The system comprises a plurality of cameras, a computer and a display screen, wherein the cameras are uniformly arranged in a region to be detected, the region to be detected is ensured to be contained in a monitoring range, the distribution of sampling points is specifically designed according to the actual condition, the actual condition of the sampling region is determined through the sampling points, the computer processes pictures according to the sampling points and feeds back whether the region to be detected changes, and the display screen displays a new image of the region to be detected, which changes.
In real life, the parking lot only displays whether vacant parking spaces exist or not, the positions of the vacant parking spaces are not displayed, and due to the fact that no special parking space guidance system exists on a public road, a plurality of vehicles continuously roam for finding the parking spaces, and urban traffic flow is too large to cause congestion.
Based on the above problems, the present embodiment provides a parking space recognition device and method based on machine vision, which implement detecting the change of the used parking space in the parking lot, and implement displaying the position of the vacant parking space in real time.
The parking space identification method based on the machine vision is based on the following steps.
Step S101: and designing sampling points on the parking places in the parking lot.
Step S102: the image is processed.
Step S103: and acquiring initial information values of the plurality of sampling points, and acquiring judgment of the background array on the state of the background array by the background array to acquire a sampling point state array.
Step S104: and judging the state of each parking space according to the state array.
Step S105: updating the background array.
Step S106: and updating the distribution map of the vacant parking spaces.
Step S101, designing sampling points on the parking lot in the parking lot comprises: and designing the position and the number of the sampling points and the area of the sampling points. Fig. 2 and 3 show a design scheme of sampling points, wherein a plurality of sampling points are arranged in a rectangular array in a parking space, and the outermost sampling points are arranged on a parking space line of the parking space; wherein, the row spacing can be set to 30cm, and the column spacing can be set to 50 cm. Of course, other designs of the sampling points can be adopted, such as a plurality of sampling points annularly arrayed in the parking space.
The rectangular parking space image usually comprises information of tens of thousands to hundreds of thousands of pixel values, and because the video image has high spatial information redundancy and the parking space detection does not care about the detailed information of the parking vehicles, a plurality of characteristic pixel points are selected in the area image of the parking space to sufficiently describe the state of the parking space. According to the principle, a plurality of characteristic pixel points with information reflecting parking space state are processed, and the characteristic pixel points are called sampling points. The state detection calculation amount of the parking space can be reduced by multiple orders of magnitude through the customization, and therefore method support is provided for realizing the parking space detection based on machine vision in the embedded system.
After the parking space and the sampling point are customized, the video detection of the state of the parking space can be realized. Step S102, processing the image, namely acquiring high-frequency regional image information of a monitoring video image through software, and performing autonomous preprocessing on the image information through machine vision, namely performing noise reduction, enhancement and the like on the image. The clearer image can be obtained through the step S102, and the more accurate judgment of the parking space image in the area is facilitated.
Where denoising the image may convert the image to the frequency domain. The image characteristics are described by taking frequency (namely wave number) as an independent variable, the spatial variation of pixel values of one image can be decomposed into linear superposition of simple vibration functions with different amplitudes, spatial frequencies and phases, the composition and distribution of various frequency components in the image are called as a spatial frequency spectrum, and the decomposition, processing and analysis of the frequency characteristics of the frequency of the image are called as frequency domain processing or wave number domain processing. The image may be converted to the spatial domain and processed, and the processing of the capsule value directly in the image space with the length (distance) as an argument is referred to as spatial domain processing. The spatial domain is also called an image space, and is a space composed of image elements. The space domain and the frequency domain can be converted to each other, and the well-established frequency domain technology can be referred to in the frequency domain, and the processing steps are generally as follows: firstly, performing two-dimensional discrete Fourier transform or wavelet transform on the image, and converting the image from an image space to a frequency domain space; and analyzing and processing the frequency spectrum of the image in the frequency domain to change the frequency characteristics of the image. I.e. the frequency spectrum of the different digital filter images is designed for filtering. The images after spatial domain processing and/or frequency domain processing eliminate random noise in the images, so that the images become smooth.
After the spatial domain and the frequency domain processing of the image, the edge and the outline of a scene in the image can be strengthened by enhancing radian contrast of the image, namely, the contrast of the detail edge of the image is increased. This processing method works to emphasize the edges and contours of the scene in the image by enhancing the gray contrast of the textured edges of the image. That is, the contrast of the detail edge of the image is increased, which helps the eyes to see the detail of the image clearly, and further enhances the recognition capability of the target.
And enhancing the contrast of the image to enhance the image. Histogram processing can be selected for the image, wherein the histogram can be classified into a gray histogram, a color histogram and a visual histogram, and the requirements for the histogram are different from the different dimensions. Aiming at the problems that the one-dimensional threshold segmentation algorithm is poor in adaptability, is easily interfered by noise and is high in complexity, the gray level histogram comprehensive threshold segmentation algorithm based on region division is provided. The algorithm utilizes pixel gray level and domain mean value to form a two-dimensional space, a one-dimensional histogram is constructed in the two-dimensional space by utilizing a region division method, then three classic segmentation algorithms of minimum error, maximum entropy and maximum between-class variance are integrated to construct a new threshold selection method, and finally the obtained threshold segmentation algorithm has stronger adaptation and steady noise resistance: meanwhile, compared with a two-dimensional algorithm, the method is more in adaptive noise variety and much smaller in algorithm calculation complexity. The simplest euclidean distance metric method is usually adopted for the conventional color distance metric, but in an HSV (Hue, Saturation, brightness) color space, because the contribution degree of each component to the color is different, the simple color distance metric method cannot be well transplanted to the space, and in the embodiment of the present invention, a gray histogram is adopted to process an image. The histogram correction enhancement technique is based on a histogram as a transformation basis, and can make the image histogram after transformation into a desired shape, thereby increasing the contrast and eliminating noise. The image processing enhances the edges and contours of the scene in the image by enhancing the gray contrast of the textured edges of the image. That is, the contrast of the detail edge of the image is increased, which helps to distinguish the detail of the image, and further enhances the recognition capability of the target.
Different digital filters are designed to filter the image, the filtering mode adopted by the embodiment of the invention is bilateral filtering, the bilateral filtering is a filter which simultaneously considers the space difference and the strength difference of the box number, the bilateral filter is carried out on a Gaussian filter, and the Gaussian filter is a Gaussian filter
Wherein, W is weight, i and j are pixel index, k is normalization constant, and from the formula of the Gaussian filter, it can be seen that the spatial distance between the pixels of the weight W carton has a relationship, so that what the content of the image has the same filtering effect; the bilateral filter is slightly modified on the original Gaussian filter, and comprises the following steps:
wherein, I is the intensity value of the pixel, and the weight can be reduced and the filtering effect can be reduced at the place with large intensity difference, so the bilateral filtering is at the area with small pixel intensity change, the bilateral filter has the same effect as the Gaussian filter, but at the place with larger intensity gradient such as the image edge, the bilateral filter can keep the original gradient.
The bilateral filter adopted in the embodiment of the invention can carry out filtering while keeping the original gradient on the sampling point.
Step S103, acquiring initial information values of the plurality of sampling points, and obtaining judgment of the background array on the state of the background array by the background array to obtain a sampling point state array. And converting the background image into an information value to obtain a background array represented by pixel points. And obtaining the state array of the sampling points by judging whether the state of the sampling points on the background array chart is covered or not.
And step S104, judging the state of each parking space according to the state array. Namely, the result of whether the parking space has the vehicle is obtained after the state array of the sampling points is obtained.
In step S105, updating the background array includes: and repeating the steps of obtaining the real-time information values of the plurality of sampling points, judging the real-time information values to obtain a state array, and judging the state of each parking space according to the state array. And acquiring images of each whole parking lot in real time, analyzing the acquisition point area in real time, and comparing the filtered background array to determine whether a large change occurs.
And step S106, updating the free parking space distribution map. And if the change is large, updating the background array, sending a processing result to a display, and receiving the result of the processor by the display to display the state of each parking space. The change is reflected to the display screen, the change of whether the parking space of the parking lot is vacant or not is updated on the display screen in real time, and the position of the vacant parking space is displayed, so that a new vehicle entering the parking lot can be quickly and quickly parked.
The embodiment of the invention adopts fewer sampling points to analyze the picture and judges the state of the parking space, thereby overcoming the defects of poor flexibility, high manufacturing cost, difficult construction and maintenance and the like of detection devices such as an induction coil, infrared rays and the like. And the problems of complex algorithm, large calculation amount and the like in the conventional computer vision method are solved, so that the algorithm can be applied to an embedded system, and the practicability of the system is enhanced.
The parking space recognition device based on the machine vision provided by the embodiment comprises a memory and a processor. The storage is used for storing computer executable instructions, and the processor is used for executing the computer executable instructions of the parking space identification method based on the machine vision provided by the embodiment.
The parking space recognition device based on the machine vision also comprises a display, and the display receives the result of the processor and displays the state of each parking space.
The above-listed detailed description is only a specific description of a possible embodiment of the present invention, and they are not intended to limit the scope of the present invention, and equivalent embodiments or modifications made without departing from the technical spirit of the present invention should be included in the scope of the present invention.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
The principle and the implementation mode of the invention are explained by applying specific examples, and the description of the above examples is only used for helping understanding the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (8)
1. The utility model provides a parking stall identification method based on machine vision which characterized in that includes:
designing sampling points on the parking places in the parking lot;
processing the image;
acquiring initial information values of the plurality of sampling points, and acquiring judgment of the background array on the state of the background array by the background array to acquire a sampling point state array;
judging the state of each parking space according to the state array;
updating the background array;
and updating the distribution map of the vacant parking spaces.
2. The machine-vision-based parking space recognition method according to claim 1, wherein after the determining of the state of the background array, the method further comprises: shape filtering the array of states.
3. The machine-vision-based parking space recognition method according to claim 2, wherein the shape filtering of the state array comprises:
converting the image to a frequency domain;
different digital filters are designed to filter the image.
4. The machine vision-based parking space recognition method according to claim 1, wherein designing a plurality of sampling points in the area image of the parking space comprises: the positions and the number of the sampling points and the areas of the sampling points are designed.
5. The machine vision-based parking space identification method according to claim 1, wherein the determining the state of each parking space according to the state array comprises: the image is processed using a method of image histograms.
6. The machine-vision-based parking space identification method according to claim 1, wherein the updating the background array comprises: and repeating the steps of obtaining the real-time information values of the plurality of sampling points, judging the real-time information values to obtain a state array, and judging the state of each parking space according to the state array.
7. The parking space recognition device based on the machine vision is characterized by comprising a memory and a processor;
the memory is to store computer-executable instructions;
the processor is configured to execute the instructions of the computer-executable method of any of claims 1-8.
8. The parking space recognition device of claim 7, further comprising a display, wherein the display receives the result of the processor and displays the status of each parking space.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010958401.3A CN112016517A (en) | 2020-09-14 | 2020-09-14 | Parking space identification method and device based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010958401.3A CN112016517A (en) | 2020-09-14 | 2020-09-14 | Parking space identification method and device based on machine vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112016517A true CN112016517A (en) | 2020-12-01 |
Family
ID=73521951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010958401.3A Pending CN112016517A (en) | 2020-09-14 | 2020-09-14 | Parking space identification method and device based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112016517A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102110376A (en) * | 2011-02-18 | 2011-06-29 | 汤一平 | Roadside parking space detection device based on computer vision |
| CN105857306A (en) * | 2016-04-14 | 2016-08-17 | 中国科学院合肥物质科学研究院 | Vehicle autonomous parking path programming method used for multiple parking scenes |
| CN106971602A (en) * | 2017-03-29 | 2017-07-21 | 深圳市金溢科技股份有限公司 | The condition detection method and laser car test equipment of a kind of parking position |
| CN107665599A (en) * | 2016-07-28 | 2018-02-06 | 北海和思科技有限公司 | The parking position automatic identifying method of view-based access control model detection |
| US20180240337A1 (en) * | 2015-10-27 | 2018-08-23 | Municipal Parking Services, Inc. | Parking space detection method and system |
-
2020
- 2020-09-14 CN CN202010958401.3A patent/CN112016517A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102110376A (en) * | 2011-02-18 | 2011-06-29 | 汤一平 | Roadside parking space detection device based on computer vision |
| US20180240337A1 (en) * | 2015-10-27 | 2018-08-23 | Municipal Parking Services, Inc. | Parking space detection method and system |
| CN105857306A (en) * | 2016-04-14 | 2016-08-17 | 中国科学院合肥物质科学研究院 | Vehicle autonomous parking path programming method used for multiple parking scenes |
| CN107665599A (en) * | 2016-07-28 | 2018-02-06 | 北海和思科技有限公司 | The parking position automatic identifying method of view-based access control model detection |
| CN106971602A (en) * | 2017-03-29 | 2017-07-21 | 深圳市金溢科技股份有限公司 | The condition detection method and laser car test equipment of a kind of parking position |
Non-Patent Citations (1)
| Title |
|---|
| 杨泷迪等: "视觉的车位智能识别研究", 《沈阳理工大学学报》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104112370B (en) | Parking lot based on monitoring image intelligent car position recognition methods and system | |
| JP4571987B2 (en) | System and method for detecting features from an image of a vehicle | |
| CN109635733B (en) | Parking lot and vehicle target detection method based on visual saliency and queue correction | |
| US20140153817A1 (en) | Patch Size Adaptation for Image Enhancement | |
| CN108399424B (en) | Point cloud classification method, intelligent terminal and storage medium | |
| CN111382658B (en) | Road traffic sign detection method in natural environment based on image gray gradient consistency | |
| CN109635722B (en) | Automatic identification method for high-resolution remote sensing image intersection | |
| CN109784171A (en) | Car damage identification method for screening images, device, readable storage medium storing program for executing and server | |
| CN104700405A (en) | Foreground detection method and system | |
| CN111652033A (en) | Lane line detection method based on OpenCV | |
| Ouma et al. | Urban features recognition and extraction from very-high resolution multi-spectral satellite imagery: a micro–macro texture determination and integration framework | |
| CN110276318A (en) | Nighttime road rains recognition methods, device, computer equipment and storage medium | |
| CN113537211A (en) | Deep learning license plate frame positioning method based on asymmetric IOU | |
| Mingzhou et al. | Detection of highway lane lines and drivable regions based on dynamic image enhancement algorithm under unfavorable vision | |
| CN110070545B (en) | Method for automatically extracting urban built-up area by urban texture feature density | |
| CN116597270A (en) | Road damage target detection method based on attention mechanism integrated learning network | |
| Kumar et al. | Traffic surveillance and speed limit violation detection system | |
| TWI498830B (en) | A method and system for license plate recognition under non-uniform illumination | |
| CN112465735A (en) | Pedestrian detection method, device and computer-readable storage medium | |
| CN112016517A (en) | Parking space identification method and device based on machine vision | |
| Dhar et al. | Interval type-2 fuzzy set and human vision based multi-scale geometric analysis for text-graphics segmentation | |
| CN110633705A (en) | Low-illumination imaging license plate recognition method and device | |
| CN115512302A (en) | Vehicle detection method and system based on improved YOLOX-s model | |
| CN114565597A (en) | Nighttime road pedestrian detection method based on YOLOv3-tiny-DB and transfer learning | |
| CN111027441B (en) | Road extraction method based on airborne hyperspectral remote sensing image |
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 |