A method and system for reconstructing complex parking scenes and automatically parking on a vehicle

Abstract

The present invention provides a method and system for on-vehicle complex parking space scene reconstruction and automatic parking. The method obtains parking space information around the vehicle in real time through the vehicle's acquisition and identification module, and uploads it to the processing module through the on-vehicle Ethernet, including parking space number, coordinates, category, attributes and angle information; the processing module restores the parking space information and reconstructs the scene, and transmits the reconstructed scene to the display system for display; when the user selects an available parking space and there are no special circumstances, parking is completed according to the set trajectory and the result is displayed. If the parking space is occupied, the user is informed and the parking process ends. This achieves a significant improvement in the automatic parking system in parking space information acquisition, complex parking space scene reconstruction and real-time display, solves the problems of two-dimensional reconstruction, data transmission and scene reconstruction real-time update delays in traditional parking systems, and greatly improves the reliability of automatic parking and user experience.

Description

Technical Field

[0001] The present invention relates to the field of automatic parking for vehicles, and in particular to a method and system for reconstructing complex parking scenarios and automatic parking on a vehicle. Background Art

[0002] With the rapid development of in-vehicle intelligence, more and more intelligent driving features are gaining popularity among users, with automatic parking being one of them. Currently, parking spaces in underground parking lots and scenic area parking lots are very tight, making parking more difficult. Existing parking features in cars combine onboard radars and cameras to transmit collected data to the cockpit controller in real time, recreating the scene on the onboard display, allowing users to observe the surrounding environment in real time. The vehicle then automatically or assists the user in parking, providing users with a new driving and parking experience.

[0003] The core of an automated parking system lies in its ability to perceive the vehicle's surroundings in real time and precisely control the vehicle to complete parking maneuvers. Parking scene reconstruction technology is crucial in this process. However, current automated parking scene reconstruction technology still has some significant shortcomings, as follows: First, most existing automated parking systems rely on two-dimensional (2D) scene reconstruction. This technology can provide basic environmental information in many situations, but it has limitations in complex parking scenarios. 2D scenes lack depth information and cannot accurately reflect spatial structure and obstacle heights, which can easily lead to the system misjudging the surrounding environment. For example, when faced with ramps, underground garages, or other scenes with significant three-dimensional spatial changes, the effectiveness of 2D reconstruction is significantly reduced, affecting parking accuracy and safety. Furthermore, most current automated parking systems rely on the Controller Area Network (CAN bus) for data transmission. While the CAN bus is widely used in the automotive industry and offers a certain degree of stability and reliability, its transmission rate and bandwidth are limited, easily creating bottlenecks when handling large amounts of sensor data (such as cameras and LiDAR). The data delay caused by this transmission method affects the system's response to real-time scene changes. Due to these reasons, existing systems cannot achieve real-time updates during scene reconstruction. The delay in data transmission and the limitation of processing power cause the system to lag in its perception of environmental changes, especially in dynamically changing parking scenarios (such as moving pedestrians or vehicles). This can easily cause the reconstructed scene information to lag behind the actual situation, increasing the risk of collision and reducing parking efficiency and safety. Summary of the Invention

[0004] In response to the deficiencies in the prior art, the purpose of the present invention is to provide a method and system for vehicle-mounted complex parking scene reconstruction and automatic parking, which achieves significant improvements in the automatic parking system in terms of parking space information acquisition, complex parking scene reconstruction and real-time display, solves the problems of two-dimensional reconstruction, data transmission delay and real-time update delay existing in traditional systems, and greatly improves the reliability of automatic parking and user experience.

[0005] In order to achieve the above technical effects, the present invention adopts the following technical solutions:

[0006] According to a first aspect of the present invention, a method for reconstructing a complex parking scene and automatically parking a vehicle is provided, comprising the following steps:

[0007] Step S1: Acquire parking space information around the current vehicle in real time from the camera system and vehicle sensors, and upload the parking space information to the intelligent cockpit domain controller CDC via the vehicle Ethernet; the real-time parking space information acquired includes parking space number ID, parking space coordinates, parking space category, parking space attributes and angle information; the parking space number ID ranges from 1 to N, and the same ID represents the same parking space. Until the parking space number reaches the upper limit N, the newly searched parking space will be reassigned 1, and this cycle will continue; the parking space coordinates are obtained in the following way: after identifying the parking space near the vehicle, the rear axle center of the vehicle is taken as the origin, the vehicle's forward direction is the X positive coordinate, and the vehicle's left is the Y positive coordinate; the acquired parking space coordinate information includes the coordinates of the corner point in front of the proximal end point and the coordinates of the corner point behind the proximal end point;

[0008] Step S2: After receiving the above parking space information, the intelligent cockpit domain controller CDC restores the information and reconstructs the scene. After the scene reconstruction is completed, the reconstructed scene is transmitted to the display system for display. When reconstructing the scene, the intelligent cockpit domain controller CDC first draws the size of the parking space at the center point of the canvas, including the length and width of the parking space, and then rotates and offsets the drawn parking space to the actual relative position. The scene reconstruction includes scene reconstruction of horizontal, vertical, diagonal parking spaces and parking spaces without parking space lines detected by radar.

[0009] Step S3: When the user selects an available parking space, if there are no special circumstances in the parking space selected by the user, the parking process is completed according to the set trajectory and the corresponding display is completed on the display system; if the parking space selected by the user is occupied by other users, the user is informed that the parking space is occupied and the parking process ends.

[0010] Preferably, the parking space attribute indicates whether the parking space is available for parking; and the parking space type includes a horizontal parking space, a vertical parking space, or an inclined parking space.

[0011] Preferably, in step S2, the smart cockpit domain controller CDC refreshes the update data 10 times per second to achieve scene reconstruction.

[0012] Preferably, in step S2, after receiving the parking space information, the intelligent cockpit domain controller CDC will first filter the received data, and ignore it if the received information exceeds a preset range.

[0013] Preferably, when the parking space drawn by the intelligent cockpit domain controller CDC in the scene reconstruction is displayed on the display system, due to the limited display area, the real-time video stream is displayed on the top of the display screen, and the scene reconstruction interface is displayed on the bottom.

[0014] Preferably, the intelligent driving domain controller ADCU identifies the nearest parking space or the parking space selected by the user, and feeds back the information to the cockpit domain controller CDC. When the cockpit domain controller CDC receives the information of the nearest parking space or the parking space selected by the user fed back by the intelligent driving domain controller ADCU, the cockpit domain controller CDC highlights the color of the parking space by default.

[0015] Preferably, during the automatic parking process, the cockpit domain controller CDC will update the real-time vehicle position and display the relative position between the current vehicle position and the parking space in real time. When parking is completed, the relative position of the vehicle in the parking space will be displayed on the display system.

[0016] According to a second aspect of the present invention, a system for reconstructing complex parking scenarios and automatically parking a vehicle is provided, which is used in the above-mentioned method for reconstructing complex parking scenarios and automatically parking a vehicle, and includes the following modules:

[0017] The acquisition and recognition module, including the camera system and vehicle sensors, is used to obtain scene images and driving information, that is, parking space information around the current vehicle in real time, and upload the acquired parking space information to the processing module via the vehicle Ethernet;

[0018] The processing module includes a smart cockpit domain controller (CDC), which is used to reconstruct the parking space scene based on the parking space information provided by the recognition module. During the parking process, the processing module updates the vehicle's position in real time and displays the relative position between the current vehicle position and the parking space in real time.

[0019] The display module includes a display screen for displaying the reconstructed parking scene. Due to the limited display area, the upper part of the display screen displays the real-time video stream, and the lower part displays the scene reconstruction interface;

[0020] The control module, including the intelligent driving domain controller ADCU, is used to identify the nearest parking space or the parking space selected by the user. When there are no special circumstances in the parking space selected by the user, the control module controls the vehicle to park in the parking space according to the set trajectory; when the parking space selected by the user is occupied by other users, the control module controls the vehicle to actively end the parking process.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] 1. The method for reconstructing complex parking scenarios and automatically parking provided by the present invention uses on-board Ethernet for data transmission, compared with traditional CAN bus transmission. This significantly improves data transmission speed and bandwidth, reduces data transmission delay, ensures real-time updating of parking space information, and thus ensures the timeliness of reconstructing complex parking scenarios.

[0023] 2. The in-vehicle method for reconstructing complex parking spaces and enabling automated parking enhances the accuracy of parking space location and classification by acquiring and uploading detailed parking space information in real time, including parking space ID, coordinates, category, attributes, and angle information. It also supports horizontal, vertical, diagonal, and line-free parking space reconstruction, adapting to diverse parking scenarios. Furthermore, this method also displays information indicating whether a parking space is available, facilitating quick user identification and selection.

[0024] 3. The vehicle-mounted complex parking scene reconstruction and automatic parking system provided by the present invention includes an acquisition and recognition module, a processing module, a display module, and a control module. Each module has a clear division of labor and works in coordination, thereby improving the flexibility and scalability of the system; and there is no need to add an additional parking module, thereby reducing vehicle R&D costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0026] Figure 1 This is a flow chart of the method for reconstructing a complex parking space scenario and automatically parking the vehicle as described in the first embodiment;

[0027] Figure 2 Schematic diagram of parking space information for the inclined train space described in the first embodiment;

[0028] Figure 3 This is a schematic diagram of parking space numbering described in the first embodiment;

[0029] Figure 4 This is a system structure diagram for complex on-board parking scene reconstruction and automatic parking described in the second embodiment. DETAILED DESCRIPTION

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Generally, the components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0031] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for protection, but merely represents selected embodiments of the present application. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments in the present application without creative work are within the scope of protection of the present application.

[0032] First embodiment

[0033] like Figure 1 As shown, this embodiment provides a method for reconstructing complex parking scenarios and automatically parking in a vehicle. The interactive systems involved in this method include: a smart cockpit domain controller (CDC), a gateway system, an intelligent driving domain controller (ADCU), a display system, a camera system, and an on-board radar. The parking scenarios involved in this method include various working conditions such as horizontal parking, vertical parking, and inclined parking. The method for reconstructing complex parking scenarios and automatically parking in a vehicle includes the following steps:

[0034] Step S1: The acquisition and recognition module, comprised of a camera system, a radar system, and vehicle sensors, captures scene images, driving information, and parking space information surrounding the vehicle in real time. It identifies available / vacant and unavailable / occupied parking spaces near the vehicle and uploads this acquired parking space information as raw scene reconstruction data to the intelligent cockpit domain controller (CDC) via the vehicle's Ethernet. This real-time parking space information surrounding the vehicle includes the parking space ID, coordinates, parking space category, parking space attributes, relative distance, and angle information. In this step, data transmission of this parking space information is accomplished via the vehicle's Ethernet, which is faster than traditional data transmission via the CAN bus, ensuring the timeliness of parking space scene reconstruction and preventing delays.

[0035] The parking space ID ranges from 1 to N. The same ID represents the same parking space. When the parking space ID reaches the upper limit N, the newly found parking space will be reassigned 1, and the cycle continues. Figure 3As shown, parking space number information is a variable array. For example, if the upper limit of parking space numbers is 18, as the vehicle moves forward, when the slot ID reaches the upper limit of 18 from 1, the next detected parking space number will be counted in a loop, that is, the parking space number of the 19th parking space will become 1, and the parking space number of the 20th parking space will be 2. Alternatively, within the range of parking spaces 1-18, when the vehicle is reversing, if the parking space initially numbered 1 is recognized again, its parking space number can still be assigned 1.

[0036] like Figure 2 As shown, the parking space coordinates are obtained as follows: after identifying a parking space near the vehicle, the center of the vehicle's rear axle is used as the origin, the vehicle's forward direction is the positive X coordinate, and the vehicle's left direction is the positive Y coordinate. The obtained parking space coordinate information includes the coordinates of the corner point in front of the proximal endpoint: NF (NF-X, NF-Y) and the coordinates of the corner point behind the proximal endpoint: NR (NR-X, NR-Y).

[0037] The parking space attribute indicates whether the parking space is available.

[0038] The parking space type is a horizontal parking space, a vertical parking space or an inclined parking space, etc.

[0039] like Figure 2 As shown in the figure, taking the parking space information of the oblique train space as an example, the camera and vehicle sensor obtain the oblique train space around the vehicle on the right side of the vehicle's forward direction. A total of 5 parking spaces are identified, including 2 occupied and 3 unoccupied parking spaces. Taking the 5th parking space (SLOT5) as an example, the displayed parking space information includes: (1) parking space ID: 5; (2) parking space position coordinates: the coordinates of the corner point NF5 in front of the near end point (NF5-X, NF5-Y), the coordinates of the corner point NR5 behind the near end point (NR5-X, NR5-Y); (3) parking space category: oblique train space (indicates that the parking spaces are arranged at a certain angle); (4) attribute: available for parking; (5) angle information: 132 degrees.

[0040] Step S2: After receiving the above parking space information, the intelligent cockpit domain controller CDC refreshes the updated data 10 times per second to realize scene reconstruction. After receiving the parking space information, the intelligent cockpit domain controller CDC will first filter and process the received data. If the received information exceeds the preset range, such as 20 meters, it will be ignored to save system resources. When reconstructing the scene, the intelligent cockpit domain controller CDC completes the drawing of the parking space based on the origin of the parking space of the vehicle. It first draws the size of the parking space at the center point of the canvas, including the length and width of the parking space, and then rotates and offsets the drawn parking space to the actual relative position. The scene reconstruction of the parking space includes the scene reconstruction of the horizontal, vertical, diagonal and radar-detected parking spaces without parking space lines.

[0041] The in-vehicle infotainment (IVI) system displays parking spaces mapped by the intelligent cockpit domain controller (CDC) during scene reconstruction on the display system. Due to limited display area, the top of the screen displays a real-time video stream, while the bottom displays the scene reconstruction interface. For example, horizontal parking spaces are limited by the reconstruction area, with a default maximum of four spaces displayed simultaneously (two on each side). Vertical parking spaces can display a maximum of 18 spaces simultaneously (nine on each side). Diagonal parking spaces can display a maximum of 16 spaces simultaneously (eight on each side). As the vehicle passes by, a new space replaces the old one by default.

[0042] Step S3: The intelligent driving domain controller ADCU identifies the nearest parking space or the parking space selected by the user, and feeds back the information to the cockpit domain controller CDC. When the cockpit domain controller CDC receives the information of the nearest parking space or the parking space selected by the user fed back by the intelligent driving domain controller ADCU, the cockpit domain controller CDC highlights the color of the parking space by default.

[0043] Once the user has selected an available parking space, provided the space is clear of obstacles, vehicles, or people, the vehicle will complete the parking process according to the pre-set trajectory and display the corresponding information on the display system. The entire scene display process is continuous and non-abrupt. During the parking process, the cockpit domain controller (CDC) updates the vehicle's position in real time, displaying the relative position between the current vehicle position and the parking space. When parking is complete, the display system displays the vehicle's relative position in the space through text, voice, and image broadcasts.

[0044] If the parking space selected by the user is occupied by another user, the acquisition and recognition module sends a notification to the cockpit domain controller (CDC), which actively ends the parking process. The user is informed that the parking space is occupied and a corresponding display is added on the display system.

[0045] Second embodiment

[0046] like Figure 4 As shown, this embodiment provides a system for reconstructing complex parking scenes and automatically parking in a vehicle, which is used to implement the method for reconstructing complex parking scenes and automatically parking in a vehicle provided in the first embodiment. The system includes the following modules:

[0047] Acquisition and recognition module: This module includes a camera system and vehicle sensors, which are used to obtain scene images, driving information and parking space information around the current vehicle in real time, and upload the acquired parking space information to the processing module via the on-board Ethernet.

[0048] Processing Module: This module includes the Intelligent Cockpit Domain Controller (CDC), which is responsible for reconstructing the parking space scene based on the parking space information provided by the recognition module. To reconstruct the parking space scene, the processing module first draws the dimensions of the parking space, including its length and width, at the center point of the canvas. Once the drawing is complete, the drawn parking space is rotated and offset to its actual relative position, and the reconstructed parking space scene is uploaded to the display module for display. Furthermore, during parking, the processing module updates the vehicle's position in real time, displaying the relative position between the current vehicle and the parking space.

[0049] Display Module: This module includes a display screen that displays the reconstructed parking space scene. Due to limited display space, the top of the screen displays the real-time video stream, while the bottom displays the scene reconstruction interface. When the ego vehicle passes by, the new parking space replaces the old one by default.

[0050] Control Module: This module includes the Intelligent Driving Domain Controller (ADCU), which is responsible for identifying the nearest parking space or the space selected by the user. If the user's selected space is normal, the control module controls the vehicle to park in the space according to the set trajectory. If the selected space is occupied by another user, the control module controls the vehicle to actively terminate the parking process.

[0051] The above describes the specific embodiments of the present invention. Based on the above description, relevant personnel can make various changes and modifications without departing from the scope of the technical concept of this invention.

Claims

1. A method for reconstructing complex parking scenes and automatically parking a vehicle, characterized in that: The following steps are involved: Step S1: Obtain parking space information around the current vehicle in real time from the camera system and vehicle sensors, and upload the parking space information to the smart cockpit domain controller CDC via the vehicle Ethernet; The parking space information obtained in real time includes parking space number ID, parking space coordinates, parking space category, parking space attributes and angle information; the parking space number ID ranges from 1 to N, and the same ID represents the same parking space. Until the parking space number reaches the upper limit N, the newly searched parking space will be reassigned to 1, and this cycle will continue; the parking space coordinates are obtained in the following manner: after identifying a parking space near the vehicle, the center of the rear axle of the vehicle is used as the origin, the direction of the vehicle's forward movement is the X positive coordinate, and the direction to the left of the vehicle is the Y positive coordinate; the obtained parking space coordinate information includes the coordinates of the corner point in front of the proximal end point and the coordinates of the corner point behind the proximal end point; Step S2: After receiving the above parking space information, the intelligent cockpit domain controller CDC restores the information and reconstructs the scene. After the scene reconstruction is completed, the reconstructed scene is transmitted to the display system for display. When reconstructing the scene, the intelligent cockpit domain controller CDC first draws the size of the parking space at the center point of the canvas, including the length and width of the parking space, and then rotates and offsets the drawn parking space to the actual relative position. The scene reconstruction includes scene reconstruction of horizontal, vertical, diagonal parking spaces and parking spaces without parking space lines detected by radar. Step S3: When the user selects an available parking space, if there are no special circumstances in the parking space selected by the user, the parking process is completed according to the set trajectory and the corresponding display is completed on the display system; if the parking space selected by the user is occupied by other users, the user is informed that the parking space is occupied and the parking process ends.

2. The method for reconstructing complex parking scenes and automatically parking a vehicle according to claim 1, characterized in that: The parking space attribute indicates whether the parking space is available for parking; the parking space type includes a horizontal parking space, a vertical parking space, or an inclined parking space.

3. The method for reconstructing complex parking scenes and automatically parking a vehicle according to claim 1, characterized in that: In step S2, the intelligent cockpit domain controller CDC refreshes the update data 10 times per second to achieve scene reconstruction.

4. The method for reconstructing complex parking scenes and automatically parking a vehicle according to claim 1, characterized in that: In step S2, after receiving the parking space information, the intelligent cockpit domain controller CDC will first filter the received data. If the received information exceeds the preset range, it will be ignored.

5. The method for reconstructing complex parking scenes and automatically parking a vehicle according to claim 1, characterized in that: When the parking space drawn by the intelligent cockpit domain controller CDC in the scene reconstruction is displayed on the display system, due to the limited display area, the real-time video stream is displayed on the top of the display screen, and the scene reconstruction interface is displayed on the bottom.

6. The method for reconstructing complex parking scenes and automatically parking a vehicle according to claim 1, characterized in that: The intelligent driving domain controller ADCU identifies the nearest parking space or the parking space selected by the user and feeds this information back to the cockpit domain controller CDC. When the cockpit domain controller CDC receives the information of the nearest parking space or the parking space selected by the user fed back by the intelligent driving domain controller ADCU, the cockpit domain controller CDC highlights the color of the parking space by default.

7. The method for reconstructing complex parking scenes and automatically parking a vehicle according to claim 1, characterized in that: During the automatic parking process, the cockpit domain controller CDC will update the real-time vehicle position and display the relative position between the current vehicle position and the parking space in real time. When parking is completed, the relative position of the vehicle in the parking space will be displayed on the display system.

8. A system for reconstructing complex parking scenarios and automatically parking a vehicle, for implementing the method for reconstructing complex parking scenarios and automatically parking a vehicle as claimed in any one of claims 1 to 7, comprising the following modules: The acquisition and recognition module, including the camera system and vehicle sensors, is used to obtain scene images and driving information, that is, parking space information around the current vehicle in real time, and upload the acquired parking space information to the processing module via the vehicle Ethernet; The processing module includes an intelligent cockpit domain controller (CDC), which is used to reconstruct the parking space scene based on the parking space information provided by the recognition module; And during the process of parking in the parking space, the processing module updates the vehicle position in real time and displays the relative position between the current vehicle position and the parking space in real time; The display module includes a display screen for displaying the reconstructed parking scene. Due to the limited display area, the upper part of the display screen displays the real-time video stream, and the lower part displays the scene reconstruction interface; The control module, including the intelligent driving domain controller ADCU, is used to identify the nearest parking space or the parking space selected by the user. When there are no special circumstances in the parking space selected by the user, the control module controls the vehicle to park in the parking space according to the set trajectory; when the parking space selected by the user is occupied by other users, the control module controls the vehicle to actively end the parking process.

Images