DYNAMIC REAL-TIME PARKING ASSISTANCE SYSTEM AND METHOD FOR IT

Abstract

The present invention provides a system (100) for dynamic real-time pre-arrival parking assistance for a vehicle (114). The system (100) comprises a satellite image database (102), a vehicle-to-everything (V2X) database (104), a vehicle infotainment system (106), and a server (108). The server (108) is configured to receive a destination from the vehicle infotainment system (106), receive satellite images from the satellite image database (102) for the received destination, identify parking spaces in the received satellite images, generate an initial parking trajectory for the identified parking spaces based on the satellite image, receive V2X data corresponding to the identified parking spaces from the V2X database (104), correct the initial parking trajectory based on the received V2X data, and transmit the corrected parking trajectory to the vehicle infotainment system (106).The system (100) offers enhanced support when parking in unfamiliar locations and in adverse weather conditions.

Description

AREA OF INVENTION

[0001] The present invention relates generally to assistance systems for parking vehicles, in particular to adaptable path planning for assisted or autonomous parking of vehicles under adverse weather conditions. BACKGROUND

[0002] Parking assistance technologies are becoming increasingly important in modern vehicles to enhance driver safety, reduce stress, and improve overall parking efficiency. Parking in unfamiliar locations can be a significant challenge for drivers, especially in adverse weather conditions. Heavy rain, in particular, can severely restrict visibility and impair the function of various vehicle sensors, making the parking process more complex and potentially more dangerous. This limited visibility affects both the driver's direct line of sight and the performance of camera-based parking assistance systems.

[0003] In heavy rain, a thick curtain of water can obscure the driver's view, making it difficult to see important visual cues such as parking lines, other vehicles, pedestrians, and obstacles. This limited visibility significantly increases the risk of collisions and makes parking more challenging. Furthermore, raindrops traveling at high speed can accumulate on the lenses of the vehicle's cameras, blurring or completely blocking their view and rendering them ineffective.

[0004] Sensor impairment extends beyond cameras. While radar and lidar systems are generally more weather-resistant, heavy rain can still cause signal scattering and absorption, reducing their effectiveness in detecting obstacles and assisting with parking maneuvers. This degradation of sensor performance affects the reliability of parking assistance systems and can lead to inaccurate instructions or failure to detect nearby objects.

[0005] The combination of limited visibility, restricted sensors, and challenging surface conditions can overwhelm the driver, leading to increased cognitive load and higher stress levels. The constant noise of rain drumming on the vehicle can be distracting and further complicate the parking maneuver. Navigating an unfamiliar location under these adverse conditions demands heightened attention and rapid decision-making, as the driver must process multiple sensory inputs simultaneously.

[0006] Conventional parking assistance systems can only partially address these challenges. Many existing systems rely heavily on visual cues and may not adequately account for the impact of severe weather conditions on sensor performance. Some systems lack the ability to adapt to rapidly changing environmental conditions, potentially providing inaccurate or unreliable guidance in heavy rain.

[0007] Patent documents such as US11170236B2, US20230138464A1, and GB2602263A have attempted to solve these problems. US6814372B1 discloses a method for identifying and evaluating on-street parking spaces using satellite and aerial imagery combined with map data. The method detects vehicles, calculates lanes on the street, and derives parking options and their availability. US20230138464A1 discloses a system for identifying and verifying parking spaces from image tiles, which uses algorithms to detect, analyze, and validate parking spaces, including those that have been obscured or incorrectly identified. GB2602263A discloses a method for identifying and classifying parking areas in geographic regions based on sequentially retrieved aerial imagery.However, these solutions do not yet fully address the problems of dynamic real-time parking assistance before a vehicle arrives.

[0008] Therefore, it is necessary to overcome the aforementioned problems. An improved parking assistance system is needed that functions effectively even in difficult weather conditions, especially heavy rain. Such a system should be able to guide the driver accurately in real time, compensate for limited visibility and impaired sensor performance, and ultimately increase the safety and efficiency of the parking process. SUBJECT OF THE INVENTION

[0009] The primary object of the present invention is to provide a system for providing dynamic real-time pre-arrival parking support for a vehicle.

[0010] Another object of the present invention is to provide real-time parking support through the integration of vehicle-to-everything (V2X) technologies, satellite data and advanced image processing algorithms.

[0011] Another object of the present invention is to ensure efficient and safe parking in unusual locations and under various weather conditions, including rain and snow.

[0012] Another object of the present invention is to identify one or more free parking spaces within the parking area and to update the corrected parking trajectory to align it with these free parking spaces.

[0013] Another object of the present invention is to provide an AI-based system that assists the user in reaching the goal by taking several factors into account. SUMMARY

[0014] According to one aspect of the present invention, a system for providing dynamic, real-time pre-arrival parking assistance for a vehicle is provided. The system comprises a satellite image database, a vehicle-to-everything (V2X) database, a vehicle infotainment system, and a server. The satellite image database is configured to provide real-time satellite imagery for the received destination. The V2X database includes V2X data provided by parked vehicles in parking lots and comprises location data including GPS coordinates, real-time traffic information, road conditions, and parking lot availability updates. The server is communicatively coupled to the satellite image database, the V2X database, and the vehicle infotainment system. The server comprises one or more processors coupled to at least one memory, wherein the memory contains machine-executable instructions.When executed by one or more processors, the instructions instruct the server to receive a destination from the vehicle's infotainment system, retrieve satellite imagery from the satellite image database for that destination, identify parking spaces within the received satellite imagery, generate an initial parking trajectory for the identified parking spaces based on the satellite imagery, retrieve V2X data corresponding to the identified parking spaces from the V2X database, correct the initial parking trajectory based on the received V2X data, and transmit the corrected parking trajectory to the vehicle's infotainment system. The corrected parking trajectory can be further refined based on the vehicle's onboard sensors while navigating along the corrected trajectory to the destination.

[0015] The initial park trajectory is generated by identifying paths within the received satellite image, connecting the identified paths to the initial park trajectory, mapping the pixel coordinates of the received satellite image to the World Geodetic System 1984 (WGS84), including latitude and longitude coordinates, and converting the WGS84 coordinates into Cartesian coordinates.

[0016] The system also includes one or more cameras to provide real-time camera images of the parking spaces installed at the destination. The server is configured to identify one or more vacant parking spaces within the parking area by applying image processing algorithms to at least one of the satellite images, camera images, or a combination thereof. Furthermore, the server is configured to update the corrected parking trajectory to align it with the one or more vacant parking spaces. The vehicle is then configured to follow the updated corrected parking trajectory and automatically park in one of the vacant spaces. The image processing algorithms include one of the following methods: deep learning, machine learning, artificial intelligence, or a combination thereof.

[0017] Furthermore, the server is configured to send a booking request to a third-party application to reserve one of the available parking spaces.

[0018] According to a further aspect of the present invention, a method for providing dynamic, real-time pre-arrival parking assistance for a vehicle is provided. The method comprises receiving a destination location from a vehicle infotainment system, receiving satellite imagery for the destination location from a satellite image database, identifying parking spaces within the received satellite imagery, generating an initial parking trajectory for the identified parking spaces based on the satellite imagery, receiving V2X data from a V2X database corresponding to the identified parking spaces, correcting the initial parking trajectory based on the received V2X data, and transmitting the corrected parking trajectory to the vehicle infotainment system.The process enables efficient and accurate parking guidance before arrival by using satellite imagery and real-time V2X data, thereby reducing the time and maneuvers associated with searching for parking in unfamiliar or crowded areas.

[0019] The present invention offers significant advantages by combining adaptable real-time path planning using satellite imagery, the creation and dynamic updating of parking trajectories prior to arrival, and the assurance of assisted / autonomous parking even in adverse weather conditions. The system utilizes V2X technologies for trajectory correction and one or more parking cameras to locate available parking spaces.

[0020] The preceding sections were given as a general introduction and are not intended to limit the scope of the following claims. The described embodiments, along with further advantages, are best understood by reference to the following detailed description in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1: Exemplary block diagram of a system for dynamic real-time parking support in pre-arrival parking support for vehicles. Fig. 2: An exemplary flowchart illustrating a procedure for dynamic real-time pre-arrival parking assistance for a vehicle. Fig. 3: An example of processing satellite images to detect parking spaces and generate trajectories. DETAILED DESCRIPTION OF THE INVENTION

[0021] Aspects of the present invention are best understood by reference to the description contained herein. All aspects described herein will be better appreciated and understood when considered in conjunction with the following descriptions. However, it should be understood that the following descriptions, while indicating preferred aspects and numerous specific details thereof, are given for illustrative purposes only and should not be treated as limitations. Changes and modifications may be made within the scope described herein without departing from the spirit and scope of the invention, and the present invention includes all such modifications.

[0022] Fig. Figure 1 shows an exemplary block diagram of a system 100 for providing dynamic real-time pre-arrival parking assistance. The system 100 comprises a satellite image database 102, a vehicle-to-everything (V2X) database 104, a vehicle infotainment system 106, a server 108, a communication network 112, and optionally a parking camera transmission 110.

[0023] The satellite image database 102 is configured to receive and store high-resolution satellite images from various geographic locations. In one embodiment, the images are acquired in real time from satellites, and the satellite image database 102 is updated, for example, every five minutes. Various satellite image providers supply the satellite image database 102 with real-time satellite images for any desired destination, which are particularly useful for identifying potential parking areas. In one embodiment, the satellite image database 102 stores the satellite images in one of the following formats: Shapefile, GeoTIFF, GeoJSON, and the like.

[0024] The V2X database 104 is an integral part of System 100, storing data on vehicle-to-everything interactions. This V2X data is crucial for trajectory correction and efficient parking. It is typically provided by parked vehicles at their parking positions. In one embodiment, the V2X data may include, but is not limited to, location data such as GPS coordinates, real-time traffic information, road conditions, and updates on the parking availability of vehicles and nearby infrastructure. V2X communication enables the exchange of data between vehicles and infrastructure, thereby improving the accuracy and reliability of System 100 in supporting parking.Furthermore, the V2X data can include vehicle speed, direction of travel, parking status, and even information about the dimensions of the parking space, helping to create a comprehensive and up-to-date picture of the parking environment.

[0025] The vehicle infotainment system 106 is an interactive interface within the vehicle 114 that allows the end user to interact with the server 108. It includes audiovisual components such as a high-resolution touchscreen display, speakers, buttons, and microphones for an enhanced user experience and the transmission of parking assistance information. The vehicle user enters their desired route via the vehicle infotainment system 106. Once the route data is entered, it is retrieved from the map and processed by the system 100. The system 100, with the assistance of the vehicle infotainment system 106, provides real-time parking support. The vehicle infotainment system 106 delivers dynamic parking assistance, including visual maps, turn-by-turn directions, and real-time updates on parking availability received from the server 108.The Infotainment System 106 also allows users to interact with the parking assistance functions, such as selecting preferred parking locations or initiating autonomous parking maneuvers.

[0026] Server 108 communicates with satellite image database 102, V2X database 104, and vehicle infotainment system 106. Server 108 comprises one or more processors coupled with at least one memory. The memory contains machine-executable instructions which, when executed by the one or more processors, cause Server 108 to perform various operations.The operations include receiving the destination location from the vehicle infotainment system 106, receiving satellite imagery from the satellite image database 102 for the received destination location, identifying parking spaces within the received satellite imagery, generating an initial parking trajectory for the identified parking spaces based on the satellite imagery, receiving V2X data corresponding to the identified parking spaces from the V2X database 104, correcting the initial parking trajectory based on the received V2X data, and transmitting the corrected parking trajectory to the vehicle infotainment system 106.

[0027] In addition, the server 108 can perform further operations, including identifying free parking spaces from the identified parking spaces, using the received satellite images, V2X data and / or camera images, further adjusting the corrected parking trajectory and transmitting it to the vehicle infotainment system 106.

[0028] The 110 parking camera transmission provides real-time camera images of parking spaces from one or more cameras installed at the destination, if available. The camera images are processed to identify available parking spaces, which can then be used to further update the corrected parking trajectory.

[0029] Network 112 enables the exchange of data between the vehicle infotainment system 106 and the server 108 of system 100. In one embodiment, Network 112 is a wireless network that uses various technologies such as cellular networks (4G, 5G), Wi-Fi, Dedicated Short-Range Communications (DSRC), or other high-frequency communication protocols to ensure reliable, real-time data transmission. Network 112 enables seamless communication, allowing for rapid updates and responsive parking assistance.

[0030] Server 108 is communicatively coupled with satellite image database 102, V2X database 104, and parking camera transmission 110. In one embodiment, this coupling can be achieved through various means, including both wired and wireless connections. Wired connections can include Ethernet, fiber optic cables, coaxial cables, or other high-speed data transmission lines. Wireless connections can utilize technologies such as Wi-Fi, cellular networks (3G, 4G, 5G), satellite communication, Bluetooth, ZigBee, or other radio frequency protocols. The choice of connection type depends on factors such as distance, data volume, speed requirements, and environmental conditions, ensuring optimal performance and reliability of System 100.

[0031] In addition, the vehicle 114 has onboard sensors for final adjustment of the corrected parking trajectory during navigation to the destination.

[0032] In Fig. Figure 2 shows an example flowchart for a procedure 200 for dynamic real-time parking assistance for a vehicle. The procedure 200 is configured so that it can be implemented by the system 100.

[0033] In step 202, the procedure 200 begins by receiving the destination location from the vehicle user via the corresponding vehicle infotainment system 106 of the vehicle 114. The destination is the area where the user wants to find a parking space and ultimately park the vehicle. The server 108 receives the destination location via the communication network 112, which connects the vehicle 114 and the server 108.

[0034] In step 204, procedure 200 includes the step of receiving the satellite images corresponding to the destination. The satellite images are received by server 108 from the satellite image database 102. The satellite images provide a top-down view of the destination, which may be a parking lot or another location where the vehicle user intends to park.

[0035] In step 206, procedure 200 continues to identify the parking spaces at the destination in the received satellite images. Server 108 processes the received satellite images to locate the parking spaces using image processing algorithms.

[0036] In step 208, procedure 200 includes the step of generating an initial trajectory using the received satellite images of the parking spaces. The initial trajectory outlines the possible path of the vehicle to reach the designated parking space. The initial trajectory is generated using trajectory algorithms, including image processing and other data processing, that are available in the prior art.

[0037] In step 210 of procedure 200, the V2X data is received for the destination location. This V2X data can include location information such as GPS coordinates, real-time traffic information, road conditions, and updates on parking availability for vehicles and nearby infrastructure. The V2X data can also include vehicle speed, direction of travel, parking status (free or occupied), and even parking space dimensions, contributing to a comprehensive and up-to-date picture of the parking environment. Furthermore, the V2X data can contain information about parking spaces previously used by the vehicle, as well as entries and exits to the parking area.

[0038] In step 212, procedure 200 includes the step of receiving real-time camera images relating to the parking area at the destination via parking camera transmission 110, if available. The camera image data may include the current status of the parking spaces (i.e., free or occupied), any obstacles, and current weather conditions that could affect visibility or the parking process.

[0039] In step 214, procedure 200 corrects the original trajectory based on the V2X data received from step 210. The correction of the initial trajectory can also incorporate the camera images from step 212, if available. Server 108 generates a corrected trajectory by adjusting the original trajectory based on the received V2X data and, optionally, the camera images. The correction process uses advanced trajectory algorithms, including image processing techniques and data fusion. The V2X data, which can contain real-time information about location data, parking availability, vehicle positions, entry and exit points, and traffic conditions, is used to refine the trajectory. If available, the camera images provide additional visual context, enabling a more precise trajectory adjustment.Server 108 uses sophisticated algorithms to integrate available V2X and image data and ensure that the corrected trajectory takes into account the most up-to-date information about the parking environment.

[0040] In one embodiment, the correction process in step 214 can use older satellite imagery of the target location when V2X data and camera images are unavailable. This approach ensures that System 100 can still provide reliable support even in scenarios with limited real-time parking assistance. Using historical satellite imagery allows System 100 to utilize previously mapped parking layouts and known structures for correction. While the older images are not as current as V2X data or real-time camera images, they still offer valuable insights into the general layout and potential parking spaces in the target area.

[0041] In step 216, the method 200 includes the step of identifying one or more vacant parking spaces using advanced image processing algorithms on the identified parking spaces. The server 108 can use at least one of the satellite imagery, V2X data, camera images, or a combination thereof. In one embodiment, the server 108 analyzes the latest satellite imagery to detect potentially vacant spaces by comparing it with real-time V2X data from nearby vehicles to confirm availability and using camera data for minute-by-minute visual verification. In another embodiment, the system 100 uses deep learning, machine learning, artificial intelligence, or a combination thereof for the satellite imagery to detect parking space boundaries and identify vacant spaces based on color, shape, and contrast differences.The V2X data provides real-time information about vehicle occupancy in the vicinity, including parking spaces that are becoming available or are occupied, in order to dynamically update the availability status of parking spaces.

[0042] If available, computer vision techniques such as deep learning, machine learning, artificial intelligence or a combination thereof are also used in the analysis of the camera images to detect the presence or absence of vehicles in individual parking spaces, to identify free parking spaces alone or to provide an additional level of verification.

[0043] In step 218, the procedure 200 continues to update the corrected trajectory to compare it with the identified available parking spaces, ensuring that the assistance system guides the vehicle 114 only to available parking spaces. The server 108 processes the information about available parking spaces determined in step 216 and updates the corrected trajectory accordingly. The update includes trajectory algorithms that update the trajectory according to the identified available parking spaces. The system 100 can prioritize certain available parking spaces based on criteria such as proximity to the destination, ease of access, or user preferences set in the vehicle infotainment system 106. The updated, corrected trajectory is calculated to minimize distance and time and improve overall parking efficiency.The dynamic updating of the corrected trajectory demonstrates the system's adaptability to changing parking conditions and significantly improves user comfort by reducing the time and maneuvers associated with searching for a parking space in unfamiliar or congested areas. The continuous refinement of the trajectory based on the latest available parking information shows that the system is capable of providing highly relevant and timely parking assistance.

[0044] In one embodiment, System 100 is configured to send a booking request to a third-party service, such as an application, to reserve one of the identified available parking spaces. After a suitable available parking space is detected, Server 108 generates a booking request containing relevant information such as the parking space's location, vehicle identification, and the desired parking duration. The request is securely transmitted to a third-party parking management system. The third-party system processes the request, confirms availability, and temporarily reserves the parking space. Upon successful reservation, a confirmation is sent back to the vehicle's infotainment system, providing the user with a confirmed parking space. Server 108 is also configured to update the corrected trajectory to reflect the reserved available parking space.Reserving parking spaces increases user comfort by ensuring the availability of parking spaces before arrival and can be particularly useful in high-demand areas or during peak hours.

[0045] In step 220, procedure 200 includes the step of transmitting the corrected trajectory to vehicle 114. Server 108 transmits the updated corrected trajectory to vehicle 114 if it is possible to detect available parking spaces, as determined in step 212. Otherwise, server 108 transmits the corrected trajectory generated in step 214 to vehicle 114. System 100 ensures that vehicle 114 receives the most suitable and up-to-date parking assistance. The transmission takes place over communication network 112 using protocols that guarantee data integrity and minimize latency.

[0046] In step 222, the procedure 200 continues with the navigation of vehicle 114 to the parking spaces. If available parking spaces are successfully identified, the transmitted updated, corrected trajectory would navigate the vehicle directly to an available space, thus optimizing the parking process. In cases where the detection of available parking spaces is not possible due to data limitations or other factors, the system 100 still provides valuable support by transmitting the corrected trajectory to improve the vehicle's approach to the parking space.

[0047] In step 224, the procedure 200 includes the adjustment of the transmitted trajectory using the vehicle's onboard sensors while approaching the identified parking spaces. This final refinement is crucial for safe and precise parking, especially in dynamic environments. The vehicle's sensors, such as cameras, ultrasonic sensors, radar, and LiDAR, continuously scan the immediate surroundings and provide real-time data on obstacles, pedestrians, other vehicles, and the dimensions of the parking space. The system 100 integrates this sensor data with the previously received trajectory and makes micro-adjustments to the transmitted trajectory as needed.The adaptable approach allows the vehicle 114 to respond to sudden changes, ensure safety, achieve precision in tight spaces and deal with real-world fluctuations, ultimately increasing user confidence and enabling autonomous parking capabilities.

[0048] In step 226, procedure 200 proceeds with parking vehicle 114. System 100 is configured to use either automatic or assisted parking, depending on the availability of detected free parking spaces. System 100 initiates automatic parking of vehicle 114 if free parking spaces were detected in the previous steps. Using the transmitted updated corrected trajectory and real-time sensor data, the vehicle's autonomous parking system takes over and maneuvers the vehicle precisely into the designated free parking space. Otherwise, if no free parking spaces are initially detected, System 100 switches to assisted parking mode. In this scenario, the vehicle's sensors continuously scan the surroundings as it navigates through the parking spaces / area.If a suitable free space is found, the vehicle notifies the driver and offers to start the automatic parking process. If the offer is accepted, the vehicle performs the parking maneuver independently, using its sensors and control units to park safely and efficiently in the newly identified space.

[0049] In Fig.Figure 3 describes an exemplary processing of satellite imagery 300 for detecting parking spaces and generating trajectories. The process begins with the acquisition of a satellite image 302. The image is a high-resolution view of a specific area, collected from various satellite imagery providers. The image is stored in file formats such as Shapefile or GeoTIFF, which can contain satellite imagery, GPS data, and text attributes. Server 108 is configured to receive the satellite image 302 from the satellite imagery database 102. After receiving the satellite image 302, Server 108 performs image processing to detect the parking spaces 304. To generate the trajectory, Server 108 is configured to first use image processing algorithms to find paths between the detected parking spaces 304.The image processing algorithms use the pixel information from the satellite image to find the paths.

[0050] Once parking spaces 304 and paths 306 have been identified, an initial trajectory 308 is created. This trajectory 308 is generated by determining the points of paths 306 from the pixel information, e.g., the midpoint between the parking spaces. The paths are then connected to form the initial trajectory 308.

[0051] The pixel coordinates of satellite image 302 are then converted into latitude and longitude coordinates by georeferencing the image. Georeferencing maps the image's coordinate system to geographic coordinates, such as, but not limited to, the World Geodetic System 1984 (WGS84). To convert trajectory points into the egocentric reference frame, server 108 is configured to convert the WGS84 coordinates to Cartesian coordinates, with the origin (0,0) set to the parking lot entrance point.

[0052] In one embodiment, the conversion from WGS84 to Cartesian coordinates is usually performed using the following equations: x=R*cos(lat)*cos(lon) y=R*cos(lat)*sin(lon) z=R*sin(lat)

[0053] Here, R is the approximate radius of the Earth (e.g. 6371 km), lat is the latitude, and lon is the longitude in radians.

[0054] Converting WGS84 to Cartesian coordinates allows System 100 to represent the trajectory in a local, vehicle-centered coordinate system, enabling more precise navigation and parking maneuvers. This egocentric frame of reference provides a more intuitive representation of the vehicle's immediate surroundings, which is crucial for accurate path planning and obstacle avoidance during the parking process.

[0055] To correct the initial trajectory, System 100 also integrates V2X data, including communication and real-time updates from other vehicles in the parking area. During the integration process, V2X data from V2X database 104 is merged to correct and optimize the parking trajectory. The V2X data can include location data such as GPS coordinates, real-time information on parking space occupancy, vehicle movements within the parking area, and potential obstacles or hazards. The original parking trajectory is corrected using the V2X data by comparing the GPS coordinates with the latitude and longitude coordinates. Furthermore, server 108 can use advanced data fusion algorithms to combine this V2X data with the original trajectory derived from satellite imagery.For example, the server can use techniques such as Kalman filtering or particle filtering to estimate the most probable and efficient path, taking into account dynamic changes in the environment.

[0056] Real-time updates from other vehicles are particularly valuable for adapting the trajectory to the current scenario, for example, for identifying entry and exit points to avoid potential conflicts. This correction increases the overall efficiency of the parking system and reduces the likelihood of congestion or blockages within the parking lot.

[0057] In one embodiment, the Server 108 can also employ machine learning algorithms to continuously improve trajectory correction based on historical V2X data and successful parking maneuvers. This adaptive approach allows the System 100 to become more accurate and responsive over time, adjusting its performance to the specific characteristics and patterns of each individual parking environment.

[0058] In one embodiment, the image processing algorithms incorporate one of the techniques of deep learning, machine learning, artificial intelligence, or a combination thereof to improve parking space detection and trajectory generation. For example, convolutional neural networks (CNNs) can be used to segment and classify parking spaces in satellite imagery. Similarly, augmented learning algorithms can be used to optimize trajectory planning by learning from successful parking maneuvers. Computer vision techniques such as edge detection and contour analysis, supplemented by AI-driven object recognition, can be used to identify parking space boundaries and obstacles. The Server 108 can implement ensemble learning methods, combining the results of multiple machine learning models to improve robustness and accuracy across various parking scenarios.Recurrent neural networks (RNNs) can be used to process temporal data from V2X communication to predict parking availability and traffic flow patterns. The various algorithms can work synergistically to create a highly adaptable and efficient parking assistance system.

[0059] In one embodiment, Server 108 comprises one or more processors, memory and storage units, and wireless communication modules. The processors may include one or more central processing units (CPUs), graphics processing units (GPUs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or a combination thereof. The processors may execute machine-executable programs stored in memory and storage units to perform the methods or instructions described herein. For example, the processors may refer to one or more GPUs and CPUs configured to perform operations such as identifying parking spaces in received satellite imagery, generating the initial parking trajectory, correcting the initial parking trajectory, and identifying available parking spaces within the parking areas.The memory and storage units can include volatile and non-volatile memory or storage devices. For example, the memory and storage units can include flash memory or storage devices such as one or more solid-state drives, dynamic random-access memory (DRAM), or synchronous dynamic random-access memory (SDRAM) such as LPDDR-SDRAM (low-power double-data-rate), and embedded multimedia controllers (eMMC). The memory and storage units can store software, firmware, data (including image data), databases, or a combination thereof. The wireless communication modules can include at least one cellular communication module, a router / gateway, a Wi-Fi communication module, a Bluetooth® communication module, or a combination thereof. The wireless communication module enables the server 108 to communicate with the vehicle 114 over a network 112.

[0060] In one embodiment, the Server 108 can contain one or more machine learning (ML) models. More specifically, the ML models could be a deep learning network, an artificial neural network, or a convolutional neural network (CNN).

[0061] The present invention provides dynamic, real-time pre-arrival parking assistance that ensures efficient and safe parking under various conditions. The system uniquely combines adaptable, real-time path planning using satellite imagery with the creation and dynamic updating of the parking trajectory before arrival, thus ensuring assisted / autonomous parking even in adverse weather conditions. The system also creates a virtual graph of parking spaces for different locations using satellite imagery, replacing a physical map for trajectory generation. This solution supports parking even in adverse weather conditions by utilizing stored information.By using V2X technology to correct the trajectory based on satellite images, the system can extrapolate and create a trajectory to all parking spaces, so that a map does not need to be created for each parking space.

[0062] The present invention offers significant advantages over conventional parking assistance systems. Unlike previous methods that rely solely on satellite or aerial imagery to detect parking spaces, the present system combines real-time satellite data with V2X data and in-vehicle sensors. This multi-layered approach provides a more comprehensive and accurate representation of the parking environment and adapts to rapid changes that cannot be detected by static images alone.

[0063] Furthermore, the system differs from conventional parking assistance technologies in its ability to generate and dynamically update parking trajectories before the vehicle arrives at its destination. Pre-arrival planning, coupled with real-time adjustments, enables more efficient navigation and reduces the time spent searching for a parking space. The integration of machine learning for predictive analytics increases the system's effectiveness and provides a level of predictive assistance previously unavailable in parking solutions.

[0064] The current solution offers significant advantages for parking assistance in adverse weather conditions by utilizing stored information. When visibility is severely impaired due to heavy rain, snow, or fog, real-time camera and sensor data can become unreliable. In such cases, the system uses previously stored satellite imagery, V2X data, and successful parking trajectories to guide the vehicle. The stored information provides a reliable basis for the location of parking spaces and optimal routes, even when current visual references are obscured. The system can compare the vehicle's current position with the stored data to ensure accurate navigation.Furthermore, the system can use historical data on how other vehicles have successfully parked under similar weather conditions to adjust its guidance, taking into account reduced traction or visibility. This dynamic, real-time parking assistance system approach ensures that parking support remains effective and safe even in challenging weather conditions where conventional visual systems might fail, thus increasing the overall reliability and usability of parking assistance across a range of environmental conditions.

[0065] These embodiments serve only to illustrate the inventive concepts contained herein. Other embodiments and modifications can be made to the compositions and processes without departing from the spirit and scope of the invention. Therefore, the scope of the present invention should not be limited to the embodiments described herein, but should be defined by the appended claims and their equivalents. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] US 11170236B2

[0007] US 20230138464A1

[0007] GB 2602263A

[0007] US 6814372B1

[0007]

Claims

[1] System (100) for providing dynamic real-time pre-arrival parking assistance for a vehicle (114), the system comprising: a satellite image database (102) configured to store satellite images; a vehicle-to-everything (V2X) database (104) configured to store V2X data; a vehicle infotainment system (106) of the vehicle (114); a server (108) that is communicatively coupled with the satellite image database (102), the V2X database (104) and the vehicle infotainment system (106), wherein the server (108) comprises one or more processors coupled with at least one memory, wherein the memory contains machine-executable instructions which, when executed by the one or more processors, cause the server to: to receive a destination from the vehicle infotainment system (106); To receive satellite images from the satellite image database (102) for the received destination; to identify parking spaces in the received satellite images; to generate an initial parking trajectory for the identified parking spaces based on the satellite image; To receive V2X data corresponding to the identified parking spaces from the V2X database (104); to correct the initial parking trajectory based on the received V2X data; and to transmit the corrected parking trajectory to the vehicle infotainment system (106). [2] System (100) according to claim 1, wherein the memory further contains machine-executable instructions which, when executed by the one or more processors, cause the server to: to identify one or more free parking spaces within the parking area; and to update the corrected parking trajectory so that it matches one or more free parking spaces; wherein the one or more free parking spaces are identified on the basis of image processing techniques, wherein the image processing techniques include at least one of the methods deep learning, machine learning, artificial intelligence or a combination thereof. [3] System (100) according to claim 2, wherein the memory further contains machine-executable instructions which, when executed by the one or more processors, cause the server to: to submit a booking request to a third-party application in order to reserve one of the available parking spaces. [4] System (100) according to claim 1, wherein the satellite image database (102) is configured to provide the real-time satellite images for the received destination. [5] System (100) according to claim 2, wherein the vehicle (114) is configured to follow the updated corrected parking trajectory and automatically park in one of the available parking spaces. [6] System (100) according to claim 1, wherein the corrected parking trajectory can be further adjusted based on onboard sensors while navigating to the destination along the corrected parking trajectory. [7] System (100) according to claim 1, wherein the initial parking trajectory is generated by: Identification of paths within the received satellite image; Connecting the identified paths to form the original park trajectory; Mapping of the pixel coordinates of the received satellite image to the World Geodetic System 1984 (WGS84), including longitude and latitude coordinates; and Conversion of WGS84 coordinates to Cartesian coordinates. [8] System (100) according to claim 7, wherein the V2X data are provided from parked vehicles in the parking lots and include location data including GPS coordinates; and wherein the initial parking trajectory is corrected using the V2X data by comparing the GPS coordinates with the latitude and longitude coordinates. [9] System (100) according to claim 2, wherein the system (100) further comprises a parking camera transmission (110) to provide real-time camera images of the parking spaces from one or more cameras installed at the destination; and wherein the image processing is applied to the real-time camera images to identify the one or more free parking spaces. [10] Method (200) for providing dynamic real-time pre-arrival parking assistance for a vehicle (114), the method comprising: Receiving a destination from a vehicle infotainment system (106); Receiving satellite images for the destination from a satellite image database (102); Identification of parking spaces on the received satellite images; Generating an initial parking trajectory for the identified parking spaces based on the satellite images; Receiving vehicle-to-everything (V2X) data from a V2X database (104) corresponding to the identified parking spaces; Correcting the initial parking trajectory based on the received V2X data; and Transferring the corrected parking trajectory to the vehicle infotainment system (106).

Images