CN111854750A - Automatic parking path selection method based on intelligent visual deep learning - Google Patents
Automatic parking path selection method based on intelligent visual deep learning Download PDFInfo
- Publication number
- CN111854750A CN111854750A CN201910346034.9A CN201910346034A CN111854750A CN 111854750 A CN111854750 A CN 111854750A CN 201910346034 A CN201910346034 A CN 201910346034A CN 111854750 A CN111854750 A CN 111854750A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- space
- preset
- deep learning
- path
- 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
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an automatic parking path selection method based on intelligent visual deep learning, which comprises the following steps of: acquiring vehicle and parking space information, planning a driving path, judging whether the path is blocked and updating the parking space information; the intelligent visual deep learning method has the advantages that the intelligent visual deep learning method is implanted into an automatic parking path planning technology, the image information right in front of the vehicle is collected in real time in the process that the vehicle runs according to the planned path, if the vehicle is blocked, the size and the position of the vehicle are judged whether to completely block the planned running path or not through comparison between the collected image and a preset image, and the running path is planned again according to the size and the position.
Description
Technical Field
The invention relates to the technical field of automatic parking, in particular to an automatic parking path selection method based on intelligent visual deep learning.
Background
Computer vision is a science for researching how to make a machine "see", and in particular, it refers to that a camera and a computer are used to replace human eyes to make machine vision of identifying, tracking and measuring target, and further make image processing, and make the result processed by computer become an image more suitable for human eye observation or transmitted to an instrument for detection. In the medical field, the computer vision technology can be applied to extract characteristic information for medical diagnosis of a patient from application images such as microscope images, X-ray images, angiographic images and the like, so that corresponding treatment measures can be taken for the patient accurately and timely; in the industrial field, the computer vision technology can be applied to manage the manufacturing process, thereby realizing the quality control of the product; in the military field, the specific position of an enemy target can be detected by applying a computer vision technology so as to carry out accurate striking; in addition, the computer vision technology has wide application in the fields of navigation, monitoring, visual special effect manufacturing and the like. However, the existing computer vision deep learning method has the problems that the result obtained by processing the image is not obvious, and the local tiny characteristic information in the source image cannot be processed.
The automatic parking path selection is to automatically plan a path for a vehicle to reach a parking space according to the parking space condition of a parking lot and the position of the vehicle, but the existing automatic parking path selection method cannot judge whether a barrier vehicle exists on a driving path. Therefore, the invention provides the combination of the intelligent visual deep learning method and the automatic vehicle parking technology, the collected picture is compared with the preset picture, the comparison result is used for guiding the vehicle driving path, whether the planned vehicle driving path is blocked or not is judged, and the driving path is judged to be replanned according to the judgment result, or the vehicle is bypassed and stopped according to the original driving path.
Disclosure of Invention
The invention aims to provide an automatic parking path selection method based on intelligent visual deep learning, which utilizes a cloud processing technology and a visual deep learning method to compare an acquired picture with a preset picture, judge whether a planned vehicle driving path is blocked or not and judge whether the driving path needs to be re-planned or not.
In order to achieve the purpose, the technical scheme of the invention is as follows: an automatic parking path selection method based on intelligent visual deep learning comprises the following steps:
Step 1, downloading and storing preset pictures of a plurality of vehicle profiles of different types in advance from a cloud space, and updating the preset pictures in real time from the cloud space;
step 2, acquiring data of a parking lot to be parked from the space through 4G/5G flow, and extracting space position information and parking space conditions of each parking space;
step 3, automatically planning a plurality of vehicle driving paths for a user to select according to the parking space position information and the parking space conditions acquired in the step 2;
step 4, when the vehicle runs according to the running path in the step 3, acquiring picture information right in front of the vehicle, and performing matching calculation with the vehicle preset picture pre-stored in the step 1;
step 5, judging whether the vehicle running path is completely blocked or not according to the calculation result of the step 4 and the space position information of the vehicle;
step 6, if the vehicle running path is completely blocked, recalculating and planning the vehicle running path; if the vehicle running path is not completely blocked, guiding the vehicle to turn and re-enter the running path after exceeding the blocking vehicle;
and 7, after the vehicle enters the parking space and is flamed out, uploading the spatial position information of the vehicle to the cloud space, and updating the parking space condition after matching calculation with the spatial position information of the parking space.
The automatic parking path selection method based on intelligent visual deep learning comprises the following steps of:
step 4.1, according to the collected information of the images in front of the vehicle, the collected outline of the blocking vehicle is decomposed into a plurality of collection nodes, and the outline of the vehicle of the preset images downloaded in the cloud space is decomposed into a plurality of preset nodes;
step 4.2, matching calculation and comparison are carried out on the collection node and the preset node which are decomposed in the step 4.1, calculation results of the collection node and the preset node are fused, a matching value is calculated, and if the matching value is higher than the preset value, the type and the size of the blocked vehicle are determined;
and 4.3, taking the geometric center of the acquisition node decomposed in the step 4.1 as the spatial position information of the blocking vehicle so as to determine the specific position of the blocking vehicle in the parking lot.
Preferably, the matching calculation method in step 4.2 is a convolutional neural network or a neural network based on a genetic algorithm.
Preferably, in the step 4, a vehicle-mounted camera or an ultrasonic radar is used to collect the image information right in front of the vehicle.
In conclusion, the beneficial effects of the invention are as follows: the invention implants the intelligent visual deep learning method into the automatic parking path planning technology, collects the picture information right in front of the vehicle in real time in the process that the vehicle runs according to the planned path, judges whether the size and the position of the vehicle completely block the planned running path or not by comparing the collected picture with a preset picture if the vehicle is blocked, and replans the running path according to the judgment.
Drawings
Fig. 1 is a flowchart of an automatic parking path selection method based on intelligent visual deep learning according to the present invention.
Detailed Description
The embodiments of the present invention are further described below with reference to the drawings.
Referring to fig. 1, an automatic parking route selection method based on intelligent visual deep learning includes the following steps:
step 1, downloading and storing preset pictures of a plurality of vehicle profiles of different types in advance from a cloud space, and updating the preset pictures in real time from the cloud space;
step 2, acquiring data of a parking lot to be parked from the space through 4G/5G flow, and extracting space position information and parking space conditions of each parking space;
step 3, automatically planning a plurality of vehicle driving paths for a user to select according to the parking space position information and the parking space conditions acquired in the step 2;
step 4, when the vehicle runs according to the running path in the step 3, acquiring picture information right in front of the vehicle, and performing matching calculation with the vehicle preset picture pre-stored in the step 1;
step 5, judging whether the vehicle running path is completely blocked or not according to the calculation result of the step 4 and the space position information of the vehicle;
step 6, if the vehicle running path is completely blocked, recalculating and planning the vehicle running path; if the vehicle running path is not completely blocked, guiding the vehicle to turn and re-enter the running path after exceeding the blocking vehicle;
And 7, after the vehicle enters the parking space and is flamed out, uploading the spatial position information of the vehicle to the cloud space, and updating the parking space condition after matching calculation with the spatial position information of the parking space.
The automatic parking path selection method based on intelligent visual deep learning comprises the following steps of:
step 4.1, according to the collected information of the images in front of the vehicle, the collected outline of the blocking vehicle is decomposed into a plurality of collection nodes, and the outline of the vehicle of the preset images downloaded in the cloud space is decomposed into a plurality of preset nodes;
step 4.2, matching calculation and comparison are carried out on the collection node and the preset node which are decomposed in the step 4.1, calculation results of the collection node and the preset node are fused, a matching value is calculated, and if the matching value is higher than the preset value, the type and the size of the blocked vehicle are determined;
and 4.3, taking the geometric center of the acquisition node decomposed in the step 4.1 as the spatial position information of the blocking vehicle so as to determine the specific position of the blocking vehicle in the parking lot.
Preferably, the matching calculation method in step 4.2 is a neural network algorithm based on a genetic algorithm, wherein the number of layers of the neural network is three, the input layer neurons include a collection node and a preset node, the number of hidden layer neurons is 11, and the output layer neurons are vehicle matching values.
Preferably, in the step 4, the vehicle-mounted camera is used for collecting the picture information right in front of the vehicle, and the vehicle-mounted camera collects the picture right in front of the vehicle at a rate of 2 frames per second so as to acquire the real-time picture information right in front of the vehicle.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all equivalent structural changes made by using the contents of the present specification, or any direct or indirect application attached to other related technical fields, are included in the scope of the present invention.
Claims (4)
1. An automatic parking path selection method based on intelligent visual deep learning is characterized by comprising the following steps:
step 1, downloading and storing preset pictures of a plurality of vehicle profiles of different types in advance from a cloud space, and updating the preset pictures in real time from the cloud space;
step 2, acquiring data of a parking lot to be parked from the space through 4G/5G flow, and extracting space position information and parking space conditions of each parking space;
step 3, automatically planning a plurality of vehicle driving paths for a user to select according to the parking space position information and the parking space conditions acquired in the step 2;
step 4, when the vehicle runs according to the running path in the step 3, acquiring picture information right in front of the vehicle, and performing matching calculation with the vehicle preset picture pre-stored in the step 1;
Step 5, judging whether the vehicle running path is completely blocked or not according to the calculation result of the step 4 and the space position information of the vehicle;
step 6, if the vehicle running path is completely blocked, recalculating and planning the vehicle running path; if the vehicle running path is not completely blocked, guiding the vehicle to turn and re-enter the running path after exceeding the blocking vehicle;
and 7, after the vehicle enters the parking space and is flamed out, uploading the spatial position information of the vehicle to the cloud space, and updating the parking space condition after matching calculation with the spatial position information of the parking space.
2. The automatic parking path selection method based on intelligent visual deep learning according to claim 1, wherein the step 4 comprises the following steps:
step 4.1, according to the collected information of the images in front of the vehicle, the collected outline of the blocking vehicle is decomposed into a plurality of collection nodes, and the outline of the vehicle of the preset images downloaded in the cloud space is decomposed into a plurality of preset nodes;
step 4.2, matching calculation and comparison are carried out on the collection node and the preset node which are decomposed in the step 4.1, calculation results of the collection node and the preset node are fused, a matching value is calculated, and if the matching value is higher than the preset value, the type and the size of the blocked vehicle are determined;
And 4.3, taking the geometric center of the acquisition node decomposed in the step 4.1 as the spatial position information of the blocking vehicle so as to determine the specific position of the blocking vehicle in the parking lot.
3. The automatic parking path selection method based on intelligent visual deep learning according to claim 1, characterized in that: the matching calculation method in the step 4.2 is a convolutional neural network or a neural network based on a genetic algorithm.
4. The automatic parking path selection method based on intelligent visual deep learning according to claim 1, characterized in that: and 4, acquiring picture information right in front of the vehicle by adopting a vehicle-mounted camera or an ultrasonic radar.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910346034.9A CN111854750A (en) | 2019-04-26 | 2019-04-26 | Automatic parking path selection method based on intelligent visual deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910346034.9A CN111854750A (en) | 2019-04-26 | 2019-04-26 | Automatic parking path selection method based on intelligent visual deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111854750A true CN111854750A (en) | 2020-10-30 |
Family
ID=72951808
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910346034.9A Pending CN111854750A (en) | 2019-04-26 | 2019-04-26 | Automatic parking path selection method based on intelligent visual deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111854750A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114347845A (en) * | 2022-03-18 | 2022-04-15 | 蔚来汽车科技(安徽)有限公司 | Method, device and medium for monitoring parking path in power change station and power change station |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105946853A (en) * | 2016-04-28 | 2016-09-21 | 中山大学 | Long-distance automatic parking system and method based on multi-sensor fusion |
| EP3343436A1 (en) * | 2016-12-30 | 2018-07-04 | Hyundai Motor Company | Automatic parking system and automatic parking method |
| CN108537284A (en) * | 2018-04-13 | 2018-09-14 | 东莞松山湖国际机器人研究院有限公司 | Posture assessment scoring method based on computer vision deep learning algorithm and system |
| CN109094556A (en) * | 2018-07-25 | 2018-12-28 | 深圳大学 | A kind of automatic parking method and system based on parking stall characteristic point |
| CN109624969A (en) * | 2018-12-24 | 2019-04-16 | 北京新能源汽车股份有限公司 | Automatic parking control method and device and electric automobile |
-
2019
- 2019-04-26 CN CN201910346034.9A patent/CN111854750A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105946853A (en) * | 2016-04-28 | 2016-09-21 | 中山大学 | Long-distance automatic parking system and method based on multi-sensor fusion |
| EP3343436A1 (en) * | 2016-12-30 | 2018-07-04 | Hyundai Motor Company | Automatic parking system and automatic parking method |
| CN108275143A (en) * | 2016-12-30 | 2018-07-13 | 现代自动车株式会社 | Automated parking system and automatic parking method |
| CN108537284A (en) * | 2018-04-13 | 2018-09-14 | 东莞松山湖国际机器人研究院有限公司 | Posture assessment scoring method based on computer vision deep learning algorithm and system |
| CN109094556A (en) * | 2018-07-25 | 2018-12-28 | 深圳大学 | A kind of automatic parking method and system based on parking stall characteristic point |
| CN109624969A (en) * | 2018-12-24 | 2019-04-16 | 北京新能源汽车股份有限公司 | Automatic parking control method and device and electric automobile |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114347845A (en) * | 2022-03-18 | 2022-04-15 | 蔚来汽车科技(安徽)有限公司 | Method, device and medium for monitoring parking path in power change station and power change station |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3542211B1 (en) | Method and device and computer program for determining a representation of a spectacle glass rim | |
| CN106485233A (en) | Drivable region detection method, device and electronic equipment | |
| KR20190088376A (en) | Data managing method, apparatus and program for machine learning | |
| CN106503653A (en) | Area marking method, device and electronic equipment | |
| CN110555845A (en) | Fundus OCT image identification method and equipment | |
| CN108363953B (en) | Pedestrian detection method and binocular monitoring equipment | |
| KR102481817B1 (en) | Method for Estimating Driver's Gaze Zone by Transfer Learning | |
| DE112011105445B4 (en) | Red-eye detection device | |
| CN106162144A (en) | A kind of visual pattern processing equipment, system and intelligent machine for overnight sight | |
| DE112009000949T5 (en) | Detection of a free driving path for a vehicle | |
| CN108171162B (en) | Crowd crowding degree detection method, device and system | |
| Pech et al. | Head tracking based glance area estimation for driver behaviour modelling during lane change execution | |
| DE112019007558B4 (en) | Cognitive function estimation device, learning device and method for estimating cognitive function | |
| CN107316332A (en) | The camera and scene relating scaling method and system of a kind of application intelligent driving | |
| US20210142149A1 (en) | Method of object re-identification | |
| CN106650635A (en) | Method and system for detecting rearview mirror viewing behavior of driver | |
| CN108320799B (en) | Image analysis and recognition method for lateral flow paper strip disease diagnosis | |
| CN111854750A (en) | Automatic parking path selection method based on intelligent visual deep learning | |
| DE102021103153A1 (en) | SYSTEM AND METHOD FOR END-TO-END PREDICTION OF TRACK DETECTION UNCERTAINTY | |
| CN111845714A (en) | Automatic parking system based on intelligent visual deep learning | |
| US20230091536A1 (en) | Camera Placement Guidance | |
| Bhagat et al. | Driver gaze fixation and pattern analysis in safety critical events | |
| CN113415292A (en) | Driving takeover capability evaluation method and electronic device | |
| Peltier et al. | A Neural Mechanism for Optic Flow Parsing in Macaque Visual Cortex | |
| Konrad et al. | Classification of Safety Driver Attention During Autonomous Vehicle Operation |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201030 |
|
| RJ01 | Rejection of invention patent application after publication |