Vehicle vacancy period evaluation method and system for parking lot

Abstract

The invention provides a vehicle vacancy period evaluation method and system for a parking lot, and is applied to the field of image data processing. According to the invention, shadow coverage detection is carried out on parking image information of a parking space, a light field disturbance gradient, a background texture shielding rate and a particle motion change rate caused by air flow are further extracted, an environment disturbance amplitude sequence can be formed, trend data can be constructed, and a fine tuning behavior of a vehicle in a parking space range can be accurately described; in the real parking process of a vehicle, regular changes such as fine adjustment of a steering wheel, slow entering and slow stopping of the posture of the vehicle and continuous convergence of a pixel-level twisted track are usually accompanied, and mostly, transient parking is linear, short-time and free of random displacement in a stable approaching mode, so that the parking accuracy of the vehicle is improved by matching a moving track with a preset parking path. Transient image changes caused by short-time parking or passing can be effectively filtered out, and a parking order is established only after a complete parking behavior closed loop is recognized.

Description

Technical Field

[0001] This invention relates to the field of image data processing, and in particular to a method and system for periodically assessing vehicle vacancy in a parking lot. Background Technology

[0002] Current parking lot vacancy detection methods mostly rely on single-frame vision or single-point sensors such as geomagnetic / ultrasound to determine "whether the space is currently occupied".

[0003] However, in real-world scenarios, parking space status exhibits characteristics such as periodicity, suddenness, short-term occupancy, and rapid turnover over time: for example, delivery vehicles may make brief stops, a vehicle may be occupied by another vehicle immediately after it leaves, and misjudgments may occur due to lighting / shadows. Single-frame judgments are prone to producing "false vacancies" or "false occupancy," affecting vehicle navigation and parking space utilization. Summary of the Invention

[0004] This invention aims to solve the problem of how to distinguish between "short-term parking" and "real parking" to avoid misjudging temporary parking as occupancy, and provides a method and system for evaluating the cycle of vehicle vacancy in parking lots.

[0005] The present invention employs the following technical means to solve the technical problem: This invention provides a method for evaluating the cycle of vehicle vacancy in a parking lot, comprising: Based on the parking terminal's pre-set camera equipment for parking spaces, the parking image information of the parking spaces is collected; Determine whether the parking image information detects that the parking space is covered by a shadow; If so, the corresponding environmental disturbance amplitude sequence is extracted from the parking image information, and trend data of the environmental disturbance amplitude sequence is constructed. Based on the trend data, the fine-tuning behavior of the vehicle within the preset range of the parking space is identified, and the movement trajectory of the vehicle is obtained. The environmental disturbance amplitude sequence specifically includes light field disturbance gradient, background texture occlusion rate and particle motion change rate. The fine-tuning behavior specifically includes the vehicle pixel-level distortion trajectory caused by steering wheel swing, the repeated black and white dot tracking path, and the rhythm of alternating movement and stillness. Determine whether the movement trajectory matches the parking path preset by the parking terminal, wherein the parking path specifically includes parking action, trajectory overlap rate and final entry angle; If a match is found, parking features in the parking image information are captured within a preset time period. Based on the parking features, a parking order for the vehicle is dynamically created. Based on the parking order, parking content for the vehicle is generated at the parking lot terminal. The parking features specifically include shadow boundaries, headlight illumination areas, and ground reflection patterns. The parking content specifically includes vehicle information, parking duration, and parking space number.

[0006] Furthermore, before the step of extracting the corresponding environmental disturbance amplitude sequence from the parking image information and constructing the trend data of the environmental disturbance amplitude sequence, the method further includes: Based on the preset camera angle of the camera device, identify the blind spot area of ​​the parking space that is blocked by other parked vehicles; Determine whether the area of ​​the camera blind spot exceeds the parking assessment threshold of the camera device for the parking space; If so, based on the preset baseline structure line of the parking space, the proportion of the camera blind spot area occupying the parking space is collected. Based on the occupancy proportion, a virtual parking image of a virtual vehicle parked in the parking space is constructed. Through the virtual parking image, the occupancy disturbance signal of the parking lot terminal on the parking space is generated. The baseline structure line specifically includes the parking space boundary line, the parking space depth line, and the vehicle front line.

[0007] Furthermore, the step of identifying the vehicle's fine-tuning behavior within the preset range of the parking space also includes: Based on the parking terminal's preset fine-tuning sensitive area for the vehicle, the vehicle posture vector corresponding to the parking image information is identified, wherein the fine-tuning sensitive area specifically includes the front and rear bumpers, the two side wheel rims and the vehicle body sidewalls; Determine whether the vehicle attitude vector exhibits a preset, continuous, small change; If so, the deformation parameters of the outer contour of the vehicle's tires in consecutive frames are extracted from the parking image information. Based on the deformation parameters, the micro-change value of the tire rotation angle is calculated. Based on the vehicle body boundary in the parking image information, the number of moving pixels of the vehicle is obtained. Based on the number of moving pixels, the micro-perturbation action of the vehicle is generated. Specifically, the micro-perturbation action includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.

[0008] Furthermore, the step of generating the vehicle's parking information at the parking terminal based on the parking order further includes: Based on the vehicle characteristics pre-collected by the parking terminal for the parking space, the departure time of the vehicle in the parking space is detected. The vehicle characteristics specifically include vehicle model, license plate data and vehicle appearance. Determine whether the departure period can be settled within a preset time limit; If not, the system identifies the cumulative number of parking spaces the vehicle has occupied in the parking lot, generates pending settlement information for the vehicle in the parking lot based on the cumulative number of parking spaces, and dynamically resets the billing period for the parking spaces based on the pending settlement information.

[0009] Furthermore, the step of determining whether the parking image information detects that the parking space is covered by a shadow also includes: Based on the shadow coverage area pre-detected by the parking terminal for the parking space, the shadow type of the shadow coverage area is identified, wherein the shadow type specifically includes static structural shadow, dynamic lighting shadow and projection change caused by light source flicker; Determine whether the proportion of the shaded area exceeds the preset proportion of the parking space; If so, then based on the shadow type, construct the shadow area expansion rate when the vehicle is parked, dynamically predict the shadow movement trend of the parking space based on the shadow area expansion rate, and mark the occlusion area of ​​the vehicle when it is parked in the parking space based on the shadow movement trend.

[0010] Furthermore, the step of determining whether the movement trajectory matches the parking path preset by the parking terminal also includes: Based on the preset parking space geometry, continuous frame images of the vehicle entering the parking space are acquired, wherein the parking space geometry specifically includes a front parking area, a vehicle body alignment area, and a rear calibration area. Determine whether the continuous frame images were acquired in segments; If so, the system identifies the vehicle's contact area with the parking space, obtains the vehicle's spatial position relative to adjacent parking spaces based on the contact area, dynamically generates the optimal parking angle for the vehicle in the parking space based on the spatial position, and sends the optimal parking angle to the vehicle through the parking lot terminal.

[0011] Furthermore, before the step of acquiring parking image information of the parking space based on the camera device preset in the parking space by the parking terminal, the method further includes: Based on the real-time location of the vehicle pre-identified by the parking terminal, an empty parking space area is obtained in the parking lot, wherein the empty parking space area is specifically an vacant parking space that has been settled. Determine whether the vehicle is parked in the empty space area; If so, then based on the parking status of the vehicle, parking discount information for the vehicle is constructed, and the parking fee for the vehicle is dynamically updated based on the parking discount information.

[0012] The present invention also provides a parking lot vacancy period assessment system, comprising: The acquisition module is used to acquire parking image information of the parking space based on the camera device preset in the parking space at the parking terminal; The judgment module is used to determine whether the parking image information detects that the parking space is covered by a shadow; The execution module is configured to, if so, extract the corresponding environmental disturbance amplitude sequence from the parking image information, construct trend data of the environmental disturbance amplitude sequence, identify the fine-tuning behavior of the vehicle within a preset range of the parking space based on the trend data, and obtain the vehicle's movement trajectory. Specifically, the environmental disturbance amplitude sequence includes light field disturbance gradient, background texture occlusion rate, and particle motion change rate. The fine-tuning behavior specifically includes the vehicle's pixel-level distortion trajectory caused by steering wheel swing, multiple repetitions of the black and white dot tracking path, and alternating rhythms of movement and stillness. The second judgment module is used to determine whether the movement trajectory matches the parking path preset by the parking terminal, wherein the parking path specifically includes parking action, trajectory overlap rate and final entry angle. The second execution module is used to capture parking features in the parking image information within a preset time period if a match is found, dynamically establish a parking order for the vehicle based on the parking features, and generate parking content for the vehicle at the parking lot terminal based on the parking order. The parking features specifically include shadow boundaries, headlight illumination areas, and ground reflection patterns, and the parking content specifically includes vehicle information, parking duration, and parking space number.

[0013] Furthermore, it also includes: The identification module is used to identify the blind spot area of ​​the parking space that is blocked by other vehicles, based on the preset camera angle of the camera device; The third judgment module is used to determine whether the area of ​​the camera blind spot exceeds the parking evaluation threshold of the camera device for the parking space; The third execution module is used to, if so, collect the occupancy ratio of the camera blind spot area to the parking space according to the preset reference structure line of the parking space, construct a virtual parking image when the virtual vehicle is parked in the parking space according to the occupancy ratio, and generate the occupancy disturbance signal of the parking lot terminal to the parking space through the virtual parking image. The reference structure line specifically includes the parking space boundary line, the parking space depth line and the vehicle front line.

[0014] Furthermore, the execution module also includes: The identification unit is used to identify the vehicle posture vector corresponding to the parking image information based on the fine-tuning sensitive area preset by the parking terminal for the vehicle. The fine-tuning sensitive area specifically includes the front and rear bumpers, the wheel rims on both sides, and the side walls of the vehicle body. The judgment unit is used to determine whether the vehicle attitude vector exhibits a preset continuous small change; The execution unit is configured to, if so, extract the deformation parameters of the outer contour of the vehicle's tires in consecutive frames from the parking image information, calculate the micro-change value of the tire rotation angle based on the deformation parameters, obtain the number of moving pixels of the vehicle based on the vehicle body boundary in the parking image information, and generate the micro-perturbation action of the vehicle through the number of moving pixels, wherein the micro-perturbation action specifically includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.

[0015] This invention provides a method and system for evaluating the cyclical availability of parking spaces, which has the following beneficial effects: This invention detects shadow coverage on parking space images and further extracts light field disturbance gradients, background texture occlusion rates, and particle motion change rates caused by airflow. This generates an environmental disturbance amplitude sequence and constructs trend data, accurately depicting the vehicle's fine-tuning behavior within the parking space. During actual parking, vehicles typically exhibit regular changes such as subtle steering wheel adjustments, gradual vehicle movement and stopping, and continuous convergence of pixel-level distorted trajectories. Short-term stops, on the other hand, are mostly linear, short-term, and random displacements without stable approach patterns. By matching the movement trajectory with a preset parking path, transient image changes caused by short-term stops or passing through can be effectively filtered out. A parking order is only established after a complete parking behavior loop is identified, thereby significantly reducing false positives for temporary stops and improving the accuracy and reliability of parking space occupancy recognition. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating an embodiment of the parking lot vacancy period assessment method of the present invention. Figure 2 This is a structural block diagram of an embodiment of the parking lot vehicle vacancy period assessment system of the present invention. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. The realization of the purpose, functional features, and advantages of the invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings.

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Reference Appendix Figure 1 A parking lot vacancy period assessment method according to an embodiment of the present invention includes: S1: Based on the camera equipment preset in the parking space at the parking terminal, collect the parking image information of the parking space; S2: Determine whether the parking image information detects that the parking space is covered by a shadow; S3: If so, extract the corresponding environmental disturbance amplitude sequence from the parking image information, construct the trend data of the environmental disturbance amplitude sequence, identify the fine-tuning behavior of the vehicle within the preset range of the parking space based on the trend data, and obtain the vehicle's movement trajectory. The environmental disturbance amplitude sequence specifically includes the light field disturbance gradient, the background texture occlusion rate, and the particle motion change rate. The fine-tuning behavior specifically includes the vehicle's pixel-level distortion trajectory caused by steering wheel swing, the repeated black and white dot tracking path, and the alternating rhythm of movement and stillness. S4: Determine whether the movement trajectory matches the parking path preset by the parking terminal, wherein the parking path specifically includes parking action, trajectory overlap rate and final entry angle; S5: If a match is found, the parking features in the parking image information are captured within a preset time period. Based on the parking features, a parking order for the vehicle is dynamically established. Based on the parking order, the parking content of the vehicle is generated at the parking lot terminal. The parking features specifically include shadow boundaries, headlight illumination areas, and ground reflection patterns. The parking content specifically includes vehicle information, parking duration, and parking space number.

[0020] In this embodiment, the system uses pre-installed and calibrated cameras in parking spaces at the parking terminal to collect parking image information. The system then determines whether the parking space is covered by shadows in these images, and executes corresponding steps accordingly. For example, if the system determines that the parking image is not covered by shadows, it considers the lighting conditions in the parking area to be stable, and the ground texture, parking space markings, and vehicle outlines to be highly visible. The system can then obtain vehicle shape and occupancy information through conventional visual recognition. The system directly performs vehicle outline extraction, vehicle boundary recognition, and license plate / vehicle volume analysis on the parking image information, and combines this with geometric judgment of the vehicle's position relative to the parking space markings to quickly determine the correct parking space location. To determine whether a vehicle has actually entered the parking space, a fast path execution mode is employed to efficiently complete vehicle detection and parking space occupancy assessment. Once a vehicle is detected to be in a stable position, the system directly proceeds to the behavior confirmation and parking order creation phases, thus avoiding unnecessary computation and improving system response speed and accuracy. For example, if the system detects shadow interference in the parking space image, it considers the lighting conditions in the parking area unstable and unable to determine the parking status through conventional visual recognition. The system then extracts the corresponding environmental disturbance amplitude sequence from the parking image information. This sequence specifically includes the light field disturbance gradient, background texture occlusion rate, and particle motion change rate. The system constructs trend data of these environmental disturbance amplitude sequences. Based on different trend data, it identifies the fine-tuning behavior of vehicles within a pre-defined parking space. This fine-tuning behavior specifically includes pixel-level distortion of the vehicle's trajectory caused by steering wheel movements, repeated black-and-white dot tracking paths, and alternating rhythms of movement and stillness, thus acquiring the vehicle's movement trajectory. When shadow interference is detected, the system does not immediately identify the vehicle; instead, it first extracts the light field disturbance gradient, background texture occlusion rate, and particle motion change rate, and constructs environmental disturbance trend data. This allows it to indirectly judge the parking space status even under unstable lighting or partial occlusion conditions, thus avoiding visual misjudgments caused by light projection, vehicle shadows, or passing vehicle occlusion, improving the robustness and stability of parking space status judgment. Qualitatively, by mapping the vehicle's motion trajectory under shadow interference to pixel-level twisted trajectories, black-and-white dot tracking paths, and dynamic-static rhythm changes, the system can identify the motion patterns of real vehicles when making parking fine-tuning from the underlying structure of the image and pixel perturbations. In this way, even if the vehicle's outer contour is covered by shadows and the body texture is blurred, the system can still identify changes in the vehicle's posture through micro-displacement data, achieving accurate judgment of the vehicle's existence, rather than relying on brightness features or contour clarity. Furthermore, by adopting a comprehensive analysis method of "movement trajectory + environmental trend + pixel fine-grained twisting" under shadow interference, the system can distinguish between short-term parking and real parking behavior, ensuring that parking behavior is only confirmed when a parking path with continuity, purpose, and convergence is detected.The system then determines whether the vehicle's trajectory matches a pre-set parking path on the parking lot terminal. The parking path specifically includes parking actions, trajectory overlap rate, and final entry angle, in order to execute corresponding steps. For example, if the system determines that the vehicle's trajectory does not match a pre-set parking path, it considers the vehicle's current behavior to lack the convergence and goal-orientation characteristic of parking behavior. This means the vehicle is merely temporarily stopping, turning around, slowly detouring, adjusting its route, or briefly observing, and is not actually performing a parking operation. The system will continue to monitor changes in the vehicle's trajectory but will mark it as "non-parking path behavior," while maintaining... The parking space is displayed as "not officially occupied," but it remains available to other vehicles and the system, thus avoiding misjudgments of parking space occupancy due to short stops. Furthermore, by combining vehicle speed, rate of change of direction, parking duration, and the presence of repeated attempts to enter the space, the system further assesses whether the behavior is likely to develop into a parking trend. For example, when the system determines that the vehicle's movement trajectory matches a pre-set parking path in the parking lot terminal, it considers the vehicle to be parking in the space. The system then captures parking features from the parking image information within a pre-set time period. These features specifically include shadow boundaries, vehicle... Based on the illumination area of ​​the headlights and the ground reflection pattern, the system dynamically creates a parking order for the vehicle. Based on this order, parking details are generated at the parking lot terminal, including vehicle information, parking duration, and parking space number. By using parking trajectory, entry angle, and parking action characteristics as criteria, the system effectively distinguishes between "real parking" and "temporary parking, short-term stopping, or detours." This avoids mis-marking of parking spaces due to short-term vehicle stops, significantly improving the accuracy of parking space usage status recognition. Simultaneously, the system captures shadow boundaries, headlight illumination areas, and ground reflection patterns within a limited time period. Visual features such as iris patterns can stably represent the parking status of vehicles. By dynamically generating parking orders based on these long-term, time-converging parking features, this order generation method, based on "continuous parking evidence," does not rely on the single-frame state of the vehicle but utilizes the stable features of continuous images. This significantly improves the reliability of parking status confirmation and reduces misjudgments caused by changes in lighting, shadow jitter, or minor vehicle adjustments. Furthermore, after generating a parking order, the system automatically forms complete parking information, including vehicle information, parking duration, and parking space number, enabling the parking lot terminal to monitor the actual occupancy of each parking space, the attributes of parked vehicles, and the status of available resources in real time.

[0021] It should be noted that the corresponding environmental disturbance amplitude sequence is extracted from the parking image information, trend data of the environmental disturbance amplitude sequence is constructed, and based on the trend data, the fine-tuning behavior of the vehicle within the preset range of the parking space is identified to obtain the vehicle's movement trajectory, specifically: After detecting that a parking space is affected by shadows, changes in lighting, or environmental disturbances, the system activates an enhanced recognition mechanism. This mechanism extracts a sequence of environmental disturbance amplitudes from the parking image information (obtained through statistical analysis of subtle brightness noise in the video, serving as a quantitative indicator of the actual degree of vehicle intervention; short-term parking vehicles typically do not fully enter the parking space, so the disturbance amplitude initially decreases rapidly and then suddenly, while actual parking tends to decrease in an orderly manner). This sequence contains dynamic parameters of the environment surrounding the vehicle, such as the light field disturbance gradient, background texture occlusion rate, and particle motion change rate. The light field disturbance gradient reflects the intensity of changes in light reflection and occlusion during vehicle parking; the background texture occlusion rate characterizes the time-varying characteristics of the proportion of ground texture occlusion by the vehicle; and the particle motion change rate reflects the movement state of tiny particles caused by air disturbances when the vehicle is parked. The system continuously monitors the time-varying trends of these disturbance parameters to construct a corresponding trend data model. After obtaining trend data, the system further identifies the vehicle's fine-tuning behavior within the preset parking space range. This identification does not rely on the overall movement of the vehicle, but is based on subtle changes at the pixel level, such as the jitter trajectory of pixels at the vehicle's edge, the outer contour deformation path caused by fine-tuning of the steering wheel, and the offset trajectory of the headlight reflection spot caused by steering. By analyzing the continuous displacement and distortion at the pixel level, the system captures characteristic behaviors such as "returning into the parking space," "micro-forward positioning," and "repeated angle adjustment" during the actual parking process, thereby inferring whether the vehicle has a genuine parking intention. For example, in a real-world scenario, if a vehicle slowly drives into a parking space and the frequency of changes in the direction of the vehicle's outline edge is approximately 0.8 times per second, while the background texture occlusion rate shows a continuous increasing trend, the light field perturbation gradient gradually converges, and the particle motion trend tends to stabilize, then the system can infer that the vehicle is gradually completing the parking maneuver. However, if the vehicle continuously slides in a straight line without significant angle changes, the light field perturbation is stable, and the background occlusion changes show small, short-term oscillations, then the system can determine that it is merely passing by, temporarily stopping, or briefly pausing for observation, rather than actually parking. Specific examples are as follows: Assuming the parking lot is located on the second basement level, with a light gray epoxy resin floor, parking space B2-053 is adjacent to a white cylinder with a diameter of 45cm on its left. The cylinder obstructs the light distribution in the parking space at different times of the day. Above the parking space is a fixed camera device with a height of 2.8 meters and a lens tilted downwards at 43°. When shooting, the upper left corner area shows obvious shadows and low brightness, while the lower right corner area is brighter due to reflected light. At 17:36:18, a silver SUV (approximately 4.72 m long and 1.88 m wide) entered the camera's field of view, initially traveling at approximately 6 km / h. At this time, the vehicle was approximately 2.3 m away from the edge of the parking space. Due to the shadow covering the rear of the vehicle, the edge outline of the vehicle in the camera image appeared blurred and discontinuous, making it impossible to determine its condition based solely on its shape. The system begins recording environmental disturbance parameters: t=17:36:18, Light field perturbation gradient: 0.46 Background texture occlusion rate: 0.12 Rate of change of particle motion: 0.33 The system determined that the vehicle was still in the process of entering and that the light disturbance was severe. The vehicle has begun its initial positioning: The vehicle stopped, and the driver began the first minor adjustment while reversing; at this time: t=17:36:24, The light field perturbation gradient decreased to 0.31. Background texture occlusion rate increased to 0.25. The rate of change of particle motion increased slightly to 0.37. Characteristics of this stage: the rear of the vehicle moves backward, causing airflow disturbances, and ground dust particles show slight dragging signs, which can be detected at the pixel level by the camera. The system detected a position shift of 14px→10px→6px in the left rear corner pixel, showing a decreasing trend; The system determined that a minor adjustment was occurring, and the vehicle was not traveling in a straight line. Vehicle positioning angle correction: When the driver turns the steering wheel 8° to the right, the SUV deviates approximately 27cm to the right. at this time: t=17:36:32, The light field perturbation gradient decreased to 0.19. The background texture occlusion rate has increased to 0.43. The rate of change of particle motion decreased to 0.22. The camera image shows: The vehicle's reversing lights reflect off the ground to create a bright band that gradually moves closer to the right white line area of ​​the parking space as the vehicle moves. The pixel changes in the rear profile of the vehicle exhibit a "zigzag offset trajectory," and the pixel disturbance trajectory detected by the system shows a zigzag shape rather than a non-linear one. The system further determines: The vehicle is performing a curved parking maneuver that conforms to parking characteristics; Vehicle finally positioned: The vehicle should stop 19cm from the rear of the parking space line. Final state sampling: t=17:36:41, The light field perturbation gradient decreased to 0.07. The background texture occlusion rate remained stable at 0.51. The rate of change of particle motion decreased to 0.04. Observe over time: The shadow boundary fits stably against the lower edge of the car body. The area illuminated by the headlights disappeared. The ground reflection pattern stopped moving; The system obtained the following results: The pixel perturbation sequence converged, and the vehicle position stabilized and stopped shaking. The final parking angle was 1.7° off from the direction of the parking space. The overlap rate reached 94% (higher than the system's set threshold of 86%). The system's final determination: The vehicle perfectly matched the preset parking path, indicating a genuine parking action. The system generates a parking order: Vehicle information: License plate number Yue B-8XXH Parking start time: 17:36:41 Parking space number: B2-053 Status marker: Parked At this point, the parking space status changes from "available" to "occupied".

[0022] In this embodiment, before step S3 of extracting the corresponding environmental disturbance amplitude sequence from the parking image information and constructing the trend data of the environmental disturbance amplitude sequence, the method further includes: S301: Based on the preset camera angle of the camera device, identify the blind spot area of ​​the parking space that is blocked by other parked vehicles; S302: Determine whether the area of ​​the camera blind spot exceeds the parking evaluation threshold of the camera device for the parking space; S303: If so, then according to the preset baseline structure line of the parking space, the proportion of the camera blind spot area occupying the parking space is collected, and according to the occupancy proportion, a virtual parking image of the virtual vehicle parked in the parking space is constructed. Through the virtual parking image, the occupancy disturbance signal of the parking lot terminal on the parking space is generated. The baseline structure line specifically includes the parking space boundary line, the parking space depth line and the vehicle front line.

[0023] In this embodiment, the system identifies the blind spot area of ​​a parking space obscured by other parked vehicles based on the pre-set camera angle of the camera device. The system then determines whether this blind spot area exceeds the pre-set parking evaluation threshold of the camera device for the parking space, and executes the corresponding steps accordingly. For example, if the system determines that the blind spot area of ​​the parking space obscured by other parked vehicles does not exceed the pre-set parking evaluation threshold, the system considers that the obstruction has not seriously affected the key identification area of ​​the parking space. That is, the main identification area used to determine whether a vehicle is parked in the parking space is still within the effective monitoring range. The system will use normal weights to identify the vehicle status in the still visible area. For example, when the rear area of ​​the parking space is obscured, the system prioritizes using the front of the parking space and the left and right marking directions for vehicle identification, while maintaining normal image acquisition and movement trajectory analysis of the parking space, and continuing to monitor identification indicators such as the movement trajectory of vehicles entering the parking space, changes in vehicle body pixels, and the rate of background texture occlusion. The system ensures that the parking space identification process is not interrupted due to minor occlusions, and marks the parking space as "identifiable" rather than "identifiable but obstructed" in the system's state logic. This allows the parking space to continue participating in parking guidance, route planning, and parking space availability display, preventing minor occlusions from causing the parking space to be incorrectly hidden or marked as unavailable in the system. For example, if the system determines that the area of ​​the camera blind spot obstructing the parking space exceeds the pre-set parking evaluation threshold of the camera equipment, the system will consider the obstruction to affect the key identification area of ​​the parking space. Based on the pre-set baseline structure lines for the parking space (including the parking space boundary line, parking space depth line, and vehicle front line), the system will collect the occupancy ratio of these camera blind spots. Based on different occupancy ratios, a virtual parking image of a virtual vehicle parked in the parking space will be constructed. This virtual parking image will then generate an occupancy disturbance signal for the parking space from the parking terminal.The system determines whether the occlusion area exceeds an evaluation threshold and introduces baseline structural lines (including parking space boundary lines, parking space depth lines, and vehicle front lines) for alternative recognition. It no longer relies on a complete image within the actual visual range but can infer based on structured coordinate relationships. In cases of severe occlusion, it simulates the vehicle's actual parking state using virtual parking images, avoiding the situation where "the camera's view is incomplete → the system cannot judge → the parking space status becomes invalid." This significantly improves the system's recognition stability and fault tolerance in complex occlusion environments. Furthermore, by calculating the proportion of blind spots and constructing virtual parking images, it can still generate images even when the actual footage is unavailable. This generates complete semantic images of parking spaces, ensuring that parking spaces remain "monitorable, analyzable, and assessable" at the system's internal data level. This means that even if a parking space is severely obstructed by another vehicle, the system doesn't need to hide or mark it as invalid in the guidance interface; instead, it can continue to be used for real-time scheduling, fundamentally improving the utilization efficiency of parking resources. Furthermore, the generated obstruction disturbance signals are used not only for current assessments but also for future risk predictions. For example, if a parking space is severely obstructed for an extended period, the system can mark it as a "high-risk obstruction space," thereby reducing its recommendation weight in subsequent parking route planning or issuing adjustment suggestions to staff.

[0024] It should be noted that the virtual parking image can be used to simulate a car parking in a blocked parking space. However, due to other vehicles parked in adjacent spaces, the camera cannot continuously assist in parking while the car is parked. If other vehicles completely block the parking space, the system can generate a blocking disturbance signal to remind the driver to pay attention to parking safety.

[0025] In this embodiment, step S3, which identifies the vehicle's fine-tuning behavior within a preset range of the parking space, further includes: S31: Based on the preset fine-tuning sensitive area of ​​the vehicle by the parking terminal, identify the vehicle posture vector corresponding to the parking image information, wherein the fine-tuning sensitive area specifically includes the front and rear bumpers, the wheel rims on both sides and the side wall of the vehicle body; S32: Determine whether the vehicle attitude vector exhibits a preset continuous small change; S33: If so, extract the deformation parameters of the outer contour of the vehicle's tires in consecutive frames from the parking image information, calculate the micro-change value of the tire rotation angle based on the deformation parameters, obtain the number of moving pixels of the vehicle based on the vehicle body boundary in the parking image information, and generate the micro-perturbation action of the vehicle through the number of moving pixels. The micro-perturbation action specifically includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.

[0026] In this embodiment, the system identifies the vehicle's attitude vector corresponding to the parking image information based on a pre-defined fine-tuning sensitive area for the vehicle at the parking terminal. This sensitive area specifically includes the front and rear bumpers, wheel rims, and sidewalls. The system then determines whether these attitude vectors exhibit a pre-defined continuous small change to execute corresponding steps. For example, if the system determines that the vehicle's attitude vector in the parking image information does not exhibit a pre-defined continuous small change, the system considers that the vehicle has not performed any obvious fine-tuning action, meaning the vehicle is in a stable parking state or has not actually entered the parking action phase. The system then directly enters the vehicle stability judgment process, comparing the current attitude vector and parking position with the parking space baseline to confirm whether the vehicle is fully in place. Simultaneously, based on the stable vehicle attitude, the system performs normal parking recording and order generation operations for the parking space. The system collects parking features and generates parking orders and parking content (vehicle information, parking duration, parking space number), and maintains continuous monitoring of vehicles. If a vehicle subsequently exhibits new posture changes, its parking or fine-tuning status is reassessed to ensure the real-time and accuracy of parking behavior judgment. For example, when the system determines that the vehicle posture vector corresponding to a certain vehicle in the parking image information shows a pre-set continuous small change, the system will consider that the vehicle has made a fine-tuning action in the parking space. The system will extract the deformation parameters of the outer contour of the vehicle's tires in consecutive frames from the parking image information, calculate the micro-change value of the tire rotation angle based on different deformation parameters, obtain the number of moving pixels of the vehicle based on the vehicle body boundary in the parking image information, and generate the vehicle's micro-perturbation action through these moving pixels. The micro-perturbation action specifically includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.By extracting the deformation parameters of the vehicle's outer tire contour in consecutive frames and combining the slight changes in tire angle with the pixel movement of the vehicle body boundary, the system can accurately identify the vehicle's subtle adjustments in a parking space, such as slow forward and backward movement, slight straightening of the front of the vehicle, or slight lateral movement. Compared with traditional methods that rely solely on the overall vehicle body contour or single-frame brightness information, this method can capture subtle changes in vehicle posture, significantly reducing misjudgments of parking status caused by undetected subtle adjustments. Furthermore, even in complex environments such as uneven lighting, shadow coverage, or occlusion by adjacent vehicles, subtle adjustments in vehicle posture can still be captured by the system through tire contour deformation and pixel-level movement. This method makes the system... Even under visual interference, the system can still determine whether a vehicle is completing a parking maneuver, thus improving the stability and reliability of parking space occupancy assessment. This avoids misclassifying short-term or temporary parking as actual parking. Furthermore, by recognizing subtle vehicle movements, the system can more accurately determine the vehicle's final parking status and dynamically adjust parking orders accordingly. Through this subtle movement information, the parking lot terminal can promptly record the time, location, and posture changes of a vehicle upon completion of parking. This enhances the accuracy and traceability of the parking management system in areas such as billing, parking guidance, inspection, and abnormal behavior monitoring, thereby improving the overall intelligent operation level of the parking lot.

[0027] It should be noted that the deformation parameters of the outer contour of the vehicle's tires in consecutive frames are extracted from the parking image information. Based on the deformation parameters, the micro-change value of the tire rotation angle is calculated. According to the vehicle body boundary in the parking image information, the number of moving pixels of the vehicle is obtained. Based on the number of moving pixels, the micro-perturbation motion of the vehicle is generated, specifically: Tire outer contour extraction: The system identifies the edge pixels of the outer contour of the vehicle tire based on continuous frame images captured by the parking space terminal camera device. The tire contour is obtained through edge detection, and then continuous frame matching is performed to generate a curve sequence of tire contour changes over time. The deformation of these contours can reflect the tire rotation and attitude adjustment during the vehicle's fine-tuning process. Tire angle micro-change calculation: The system maps the tire profile change parameters in consecutive frames to the tire angle micro-change value (Δθ). The calculation method can be based on profile curve rotation fitting or least squares straight line method to fit the tire outer profile rotation angle to obtain the small changes in the tire rotation direction and amplitude. By analyzing the time series of Δθ, it can be determined whether the vehicle is making slow direction adjustments or fine-tuning reversing actions. Vehicle body boundary pixel movement extraction: The system uses the vehicle body contour boundary in consecutive frame images to calculate the pixel-level movement (Δx, Δy) of the vehicle as a whole in the X and Y directions, which is used to determine the vehicle's front-to-back or lateral micro-movements; by combining pixel displacement with tire rotation angle changes for analysis, the vehicle's fine-tuning behavior can be more accurately reconstructed. Micro-perturbation action generation: Based on the slight changes in tire angle and vehicle pixel displacement, the system abstracts the vehicle's fine-tuning actions into identifiable behavior types, such as: slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement (the vehicle slightly translates to adjust the angle). This micro-perturbation action can be used to determine whether the vehicle is in a true parking state, assisting in the generation of parking orders and confirmation of parking space occupancy. Specific example 1 is as follows: Suppose that in parking space B2-053 of an underground parking garage, a camera captures a series of frames showing a silver SUV entering the parking space: t=17:36:20: The rear of the vehicle is close to the rear line of the parking space. The tire profile is initially identified as elliptical, and the tire turning angle Δθ=0°. t=17:36:24: The vehicle slowly reverses, the left rear wheel profile rotates clockwise by Δθ=2.3°, the right front wheel profile rotates slightly counterclockwise by Δθ=-1.8°, and the vehicle moves backward by Δy=12px and moves slightly to the right by Δx=3px. t=17:36:28: The vehicle steering wheel is slightly adjusted, the tire angle Δθ changes by 1.2°, and the vehicle body moves slightly by Δy=5px and Δx=1px. The system synthesizes these continuous changes into a perturbation action sequence, which is identified as "slow reversing + slight straightening of the front of the car + slight lateral movement", and judges that the vehicle is actually making a fine-tuning and positioning. In this way, the system can not only capture the overall movement of the vehicle, but also accurately identify tire rotation and fine-tuning movements, making the recognition of perturbation behavior accurate and reliable even under conditions of changing lighting, shadow coverage, or occlusion by neighboring vehicles. Specific example 2 is as follows: Suppose that in the B2 level of the underground parking garage, there are two white sedans parked on the left and right sides of parking space B2-056. The camera is installed at a 45° angle downwards from above the parking space, and the shooting angle is partially blocked by the adjacent car. There are LED lights on the ceiling of the parking space area, which creates local shadows with alternating light and dark on the ground. At the same time, the front right corner of the parking space is blocked by a pillar, creating a local blind spot. At this moment, a dark blue SUV slowly drives into the parking space. The system performs tire contour extraction in consecutive frame images: t=10:15:02: The rear of the vehicle enters the camera's field of view. The right rear wheel is obscured by about 18% of its area by the adjacent vehicle. The left rear wheel can be fully identified. The initial tire turning angle Δθ=0°. t=10:15:06: The vehicle begins to make minor adjustments while reversing. The outline of the left rear wheel appears to rotate clockwise by Δθ=2.5°, while the outline of the right rear wheel is partially obscured. The system uses the known tire geometry and symmetry algorithm to estimate that Δθ≈2.2°. At the same time, the vehicle body pixel movement is Δy=10px (rearward movement) and Δx=2px (minor lateral movement). t=10:15:10: The vehicle performs a slight straightening motion, with the tire profile rotating Δθ by 1.3° for the left wheel and 1.1° for the right wheel, and the vehicle body slightly adjusting to the right by Δx=4px; due to the flickering of the LED lights and changes in the ground shadow, the brightness of some pixels is unstable, but the system can still accurately identify the fine-tuning motion by eliminating abnormal brightness disturbances through trend analysis. t=10:15:15: The rear of the vehicle approaches the rear line of the parking space, and the tire rotation amplitude gradually decreases Δθ≈0.5°. The vehicle body pixel moves Δy=3px and Δx=1px. The system combines the tire rotation angle and vehicle body pixel movement of consecutive frames to generate a micro-perturbation action sequence: "slow reversing + slight straightening of the front of the vehicle + slight side movement", which determines that the vehicle is making a real positioning fine adjustment. This method allows the system to accurately identify vehicle fine-tuning actions even when some tires are obscured by neighboring vehicles or when there is shadow interference on the ground. It can do so by using tire outer contour deformation parameters, tire cornering angle micro-change values, and vehicle pixel movement. This avoids misjudgment of parking status due to obstruction or changes in lighting, and improves the reliability and stability of parking space occupancy judgment.

[0028] In this embodiment, step S5, which generates the parking information of the vehicle at the parking lot terminal based on the parking order, further includes: S51: Based on the vehicle characteristics pre-collected by the parking lot terminal for the parking space, detect the time period during which the vehicle leaves the parking space, wherein the vehicle characteristics specifically include vehicle model, license plate data and vehicle appearance; S52: Determine whether the departure period can be settled within a preset time period; S53: If not, identify the cumulative number of parking spaces the vehicle has occupied in the parking lot, generate the pending settlement information for the vehicle in the parking lot based on the cumulative number of parking spaces, and dynamically reset the billing period of the parking space based on the pending settlement information.

[0029] In this embodiment, the system detects the departure time of a vehicle in a parking space based on vehicle characteristics pre-collected by the parking lot terminal. These characteristics include vehicle model, license plate data, and vehicle appearance. The system then determines whether the departure time can be settled within a pre-set duration and executes the corresponding steps accordingly. For example, if the system determines that a vehicle can be settled within a pre-set duration after leaving the parking space, it considers the vehicle's departure behavior to comply with the parking lot's billing and settlement rules. That is, the parking time is within the system's preset settlement period, the vehicle has not exceeded the time limit, or there is no abnormal departure behavior. The system then initiates the parking settlement process based on the vehicle's parking time in the parking space, the pre-collected vehicle characteristics, and the pre-collected vehicle characteristics. The system calculates the amount due based on fee rules and discount policies, and records information such as vehicle departure time, parking duration, settlement amount, and parking space number to the parking management terminal, generating a complete departure order. This order is linked to the vehicle's terminal to ensure subsequent data traceability. After a vehicle completes settlement upon departure, the system immediately updates the corresponding parking space status to "available" for subsequent vehicles to use. It also sends parking space availability information to the parking guidance system or management terminal, achieving dynamic management and efficient utilization of parking resources. For example, if the system determines that a vehicle has left a parking space but has not settled its departure within the pre-set time, the system will consider that the vehicle lingered in the parking lot after leaving the parking space, possibly due to another vehicle being used. When a vehicle parks in a designated spot, the system identifies the vehicle's cumulative parking space count within the parking lot. Based on this count, it generates pending settlement information for the vehicle and dynamically resets the billing period for each parking space based on the specific pending settlement information. The system can also identify continuous lingering behavior in the parking lot by determining if a vehicle fails to complete settlement within a preset time after leaving a parking space. This includes situations such as changing parking spaces or short stops. Based on the cumulative parking space count, it generates pending settlement information to ensure that every parking activity is recorded and billed. This effectively prevents missed parking fees due to short-term vehicle movement or changing spaces, improving the completeness and accuracy of parking lot settlements. Furthermore, it dynamically resets the billing period for each parking space based on different pending settlement information. By setting a billing period for parking spaces, the continuous parking behavior of the same vehicle in the parking lot can be reasonably split or combined for billing, ensuring that the billing logic is consistent with the actual parking behavior. This method can handle complex situations such as vehicles changing parking spaces multiple times or making short stops, improving the billing flexibility and intelligent management capabilities of the parking lot system. Furthermore, by identifying the cumulative number of parking spaces a vehicle has parked in the parking lot and generating dynamic pending settlement information, the system can update the usage status and availability of each parking space in real time, avoiding resource waste caused by long-term unsettled parking spaces. This mechanism not only ensures the accuracy of parking space status information, but also provides reliable data for guiding parking, reserving parking spaces, and parking space scheduling, improving the overall operational efficiency and user experience of the parking lot.

[0030] In this embodiment, step S2, which determines whether the parking image information detects that the parking space is covered by a shadow, further includes: S21: Based on the shadow coverage area pre-detected by the parking terminal for the parking space, identify the shadow type of the shadow coverage area, wherein the shadow type specifically includes static structural shadow, dynamic lighting shadow and projection change caused by light source flicker; S22: Determine whether the proportion of the shaded area exceeds the preset proportion of the parking space; S23: If so, then based on the shadow type, construct the shadow area expansion rate when the vehicle is parked, dynamically predict the shadow movement trend of the parking space based on the shadow area expansion rate, and mark the occlusion area of ​​the vehicle when it is parked in the parking space based on the shadow movement trend.

[0031] In this embodiment, the system identifies the shadow types of parking spaces based on the shadow coverage areas pre-detected by the parking terminal. Specific shadow types include static structural shadows, dynamic lighting shadows, and projection changes caused by light source flicker. The system then determines whether the proportion of these shadow coverage areas exceeds a pre-set upper limit for the parking space, and executes corresponding steps accordingly. For example, if the system determines that the proportion of the shadow coverage area of ​​a parking space does not exceed the pre-set upper limit, the system considers the lighting conditions of that parking space to be within an acceptable range, and the shadows have a minimal impact on vehicle recognition and parking status assessment. The system will then continue to use the standard visual recognition process based on vehicle outlines, parking space markings, shadow boundaries, etc. The system determines whether a parking space is occupied by capturing vehicle movement trajectories, posture changes, and parking characteristics in consecutive frames. It monitors vehicle entry and fine-tuning behavior, and uses shadow information to remove trends in pixel brightness changes, ensuring accuracy. Since the shadow does not substantially affect key recognition areas, the system marks the parking space as "identifiable and available," allowing it to participate in parking guidance, route planning, and management scheduling, ensuring efficient use of parking resources. For example, if the system determines that the proportion of a parking space's shadow coverage exceeds the pre-set limit, it considers the lighting conditions of that parking space excessively obscured by shadows, affecting the camera's judgment of the parking status. The system will then adjust its approach based on different shadow types. The system constructs the shadow area expansion rate when the vehicle is parked. Based on these expansion rates, it dynamically predicts the shadow movement trend of the parking space and marks the occluded area of ​​the vehicle when parked. The system can identify severely affected lighting conditions in the parking space when the shadow coverage ratio exceeds a preset threshold. For different types of shadows, the system constructs the shadow area expansion rate when the vehicle is parked. This expansion rate quantifies the degree of occlusion of the vehicle outline and key areas of the parking space, providing a quantifiable reference for subsequent vehicle recognition. This reduces the risk of misjudgment caused by lighting interference or shadow coverage. Furthermore, by constructing shadow movement trends using the shadow area expansion rate, the system can predict potential occlusion changes during vehicle parking. The system tracks the movement and expansion of shadow boundaries over time. Based on this trend, the system can mark areas that may be obscured when a vehicle is parked in real time, providing compensation information for fine-tuning action analysis, tire posture recognition, and vehicle body contour detection. This ensures the stability and reliability of vehicle parking status judgment even under complex lighting conditions. Furthermore, through quantitative and dynamic analysis of shadow coverage areas and occupancy trends, the system can accurately determine the vehicle's parking and fine-tuning status in parking spaces with complex lighting conditions, ensuring real-time updates of parking space occupancy status. The generated occupancy area information can be used for parking guidance, route planning, and abnormal parking monitoring, enabling the parking management system to operate efficiently and intelligently in complex environments, improving the utilization efficiency of parking resources and user experience.

[0032] It should be noted that when the system marks the obstructed areas, and after the parking terminal confirms that the vehicle needs to be parked, the system will communicate remotely with the vehicle in advance to inform the driver that there are these obstructed areas in the parking space. The parking terminal cannot assist the driver in parking through camera equipment. The driver needs to pay attention to parking safety and avoid the vehicle being scratched due to shadows when parking.

[0033] It should be added that, based on the shadow type, a shadow area expansion rate is constructed when the vehicle is parked. Based on this shadow area expansion rate, the shadow movement trend of the parking space is dynamically predicted. Using this shadow movement trend, the occlusion area of ​​the vehicle when parked in the parking space is marked. Specifically: The system first classifies the shadows in the parking space area, including static structural shadows (such as projections from pillars or walls), dynamic lighting shadows (such as shadow changes caused by the movement of sunlight or indoor lights), and light source flickering projections (such as rapid changes in brightness caused by the flickering of lights). For each shadow type, the system extracts the area of ​​the vehicle-occupied area covered by shadows in consecutive frame images and calculates the ratio with the actual area occupied by the vehicle to obtain the shadow area expansion rate. The expansion rate can quantify the degree of occlusion of the shadow on the vehicle outline and key identification areas of the parking space, providing numerical basis for subsequent dynamic analysis. The system then uses the time sequence of shadow area expansion rate changes in consecutive frames to establish a shadow movement trend model. By analyzing the rate, direction and periodicity of the expansion rate change, the system can predict the dynamic changes of shadows during vehicle parking, such as the expansion, contraction or directional movement of shadow boundaries. This trend analysis not only reflects the real-time changes in lighting conditions, but also helps to determine the actual impact of vehicle fine-tuning actions on shadows. Finally, based on the shadow movement trend, the system generates dynamic occlusion area markers on the vehicle outline and parking area, indicating the parts of the vehicle that may be covered by shadows at different times. For example, the rear or front of the vehicle may be partially occluded in the shadow edge area. The system marks this area as an occluded area in the image or data model for compensation processing of vehicle outline detection, fine-tuning action recognition and parking status judgment, thereby improving the accuracy of parking judgment in complex lighting environments. Specific examples are as follows: Assuming it's on level B2 of the underground parking garage, parking space B2-058 is right next to a column on the left, the overhead lights are flickering slightly, and there are noticeable shadows on the ground; When a black SUV enters the parking space, the system recognizes static structural shadows (the projection of a pillar covers the front left corner of the parking space), dynamic lighting shadows (slight movement of LED lights creates changes in the rear shadow), and light source flashing projections (flickering of the overhead lights causes alternating light and dark on the ground). The system measured the change in the proportion of the vehicle's rear area covered by shadow to the total area of ​​the rear of the vehicle over time in consecutive frames, from 0.18 to 0.27, and calculated the shadow area expansion rate. Based on the trend of the expansion rate, the system predicted that the shadow would expand to the right along the left edge of the parking space by about 12cm, and generated a dynamic occlusion area marker to cover the rear of the vehicle and the left rear wheel area. In the subsequent fine-tuning action recognition process, the system uses the information of the occluded area to compensate for the movement of tire contour and vehicle body boundary pixels, ensuring that even if the shadow occludes part of the area, it can accurately judge the vehicle's positioning and fine-tuning actions, and avoid misjudging the parking status.

[0034] In this embodiment, step S4, which determines whether the movement trajectory matches the parking path preset by the parking terminal, further includes: S41: Based on the preset parking space geometry, acquire continuous frame images of the vehicle entering the parking space, wherein the parking space geometry specifically includes a front parking area, a vehicle body alignment area, and a rear calibration area. S42: Determine whether the continuous frame image is acquired in segments; S43: If so, identify the vehicle's contact area with the parking space, obtain the vehicle's spatial position relative to adjacent parking spaces based on the contact area, dynamically generate the optimal parking angle for the vehicle in the parking space based on the spatial position, and send the optimal parking angle to the vehicle through the parking lot terminal.

[0035] In this embodiment, the system acquires continuous frame images of a vehicle entering a parking space based on a pre-set parking space geometry, specifically including a front entry area, a body alignment area, and a rear calibration area. The system then determines whether these continuous frame images were acquired in segments to execute corresponding steps. For example, if the system determines that the continuous frame images of a vehicle entering a parking space were not acquired in segments, the system assumes that the vehicle did not exhibit any obvious staged fine-tuning movements during the entire entry process. The vehicle is considered to have been parked directly in the parking space without any directional adjustments or forward / backward movement. The system will directly match the image with the parking space geometry without needing to analyze the fine-tuning trajectory or tire posture changes, quickly confirming the vehicle's position. The vehicle has been correctly positioned, and its parking status has been recorded. Upon completion of the parking process, the system directly generates a parking order, including vehicle information, entry time, parking space number, and parking status, and marks the parking space as "occupied." This allows the parking management system to update parking space status in real time, ensuring smooth operation of parking guidance, route planning, and billing processes. For example, if the system determines that consecutive frame images of a vehicle entering a parking space are segmented, it assumes the vehicle is adjusting its parking space to avoid improper parking. The system identifies the vehicle's contact area with the parking space and, based on different contact areas, obtains the vehicle's spatial position relative to adjacent parking spaces. The system dynamically generates the optimal parking angle for the vehicle within the parking space and sends this angle to the vehicle via a pre-established communication channel between the vehicle and the parking lot terminal. By analyzing segmented, continuous frame images of the vehicle within the parking space, the system can determine if the vehicle has made minor adjustments during the parking process to adapt to the available space. Based on the relationship between the vehicle and the contact area of ​​the parking space, the system calculates the vehicle's spatial position relative to adjacent parking spaces using camera equipment, dynamically generating the optimal parking angle. This guides the vehicle to park precisely within a limited space, reducing the risk of scratches, misalignment, or occupying multiple parking spaces due to improper parking angles. Simultaneously, the system sends the optimal parking angle to the vehicle via a pre-established communication channel, ensuring the vehicle... The system allows vehicles to adjust their direction and parking strategy in real time. This information feedback mechanism provides intelligent assistance during the actual parking process, not only improving the convenience of driver operation but also enhancing the parking lot's real-time control capabilities in complex environments (such as adjacent vehicles, narrow parking spaces, or obstructions from pillars). This makes parking behavior more standardized and efficient. Furthermore, by dynamically calculating the spatial position of the vehicle relative to adjacent parking spaces and generating the optimal parking angle, the system ensures that each vehicle is parked in the center of the parking space at a suitable angle, maximizing parking space utilization efficiency and reducing waste of adjacent spaces caused by vehicle misalignment. By precisely guiding vehicles to fine-tune their parking, the system can reduce the risk of vehicle collisions and the incidence of parking accidents, thereby improving the overall management level and operational safety of the parking lot.

[0036] In this embodiment, before step S1, which involves acquiring parking image information of a parking space using a camera device pre-set on the parking space by the parking terminal, the method further includes: S101: Based on the real-time location of the vehicle pre-identified by the parking terminal, obtain the empty parking space area in the parking lot, wherein the empty parking space area is specifically an vacant parking space that has been settled. S102: Determine whether the vehicle is parked in the empty space area; S103: If so, then based on the parking status of the vehicle, construct the parking discount information for the vehicle, and dynamically update the parking fee for the vehicle based on the parking discount information.

[0037] In this embodiment, the system obtains vacant parking spaces within the parking lot based on the real-time location of the vehicle pre-identified by the parking lot terminal. Specifically, vacant parking spaces are those that are empty and have been settled. The system then determines whether a vehicle is parked within a vacant parking space and executes the corresponding steps accordingly. For example, if the system determines that a vehicle has not parked in the vacant parking space assigned to it by the parking lot terminal after entering the lot, the system considers that the vehicle has not parked according to the parking lot terminal's scheduling instructions. This may indicate arbitrary parking, illegal parking, disorderly occupation of spaces, or temporary parking. The system will then re-mark the vehicle's current actual parking location based on the vehicle's real-time location trajectory and update the status of the corresponding parking space. The status is updated to "abnormal occupancy" to prevent the parking space from being misled into being "vacant and available." Simultaneously, a notification message is automatically generated indicating off-center parking or parking outside designated areas. This message is sent to the parking lot's reminder terminal via license plate association or pushed to the driver through a pre-set communication channel, reminding the driver that their vehicle is not parked in the designated space. Furthermore, the parking resources are reassessed based on the vehicle's actual parking location. The vacant area originally allocated to that vehicle is re-released as available, and guidance information for other entering vehicles is dynamically updated to ensure the rational use of parking resources and prevent confusion caused by the vehicle leading to vacancy or occupancy of existing spaces. For example, when the system detects a vehicle parking outside designated areas after entering the parking lot... When a vehicle is parked in an empty space assigned by the parking terminal, the system considers it to be parked in accordance with the terminal's dispatch instructions. Based on the vehicle's parking status, the system constructs parking discount information and dynamically updates the parking fee accordingly. By recognizing that a vehicle is indeed parked in a system-assigned empty space, the system rewards the driver for correctly following the parking terminal's dispatch instructions. By incentivizing standardized parking behavior through parking discount information, this mechanism effectively guides vehicles to follow parking guidelines, thereby reducing random parking, occupying emergency areas, or parking out of bounds, thus creating an orderly parking environment throughout the parking lot. Parking fees are dynamically updated based on parking status. A reward mechanism directly links parking costs to the degree of compliance with parking rules. This dynamic billing model not only improves the parking experience but also enhances the interaction between drivers and the parking system, making users more willing to cooperate and follow guidance, thus increasing user satisfaction and the self-discipline of parking behavior. Furthermore, by incentivizing vehicles to prioritize parking in system-assigned empty spaces, the system can maximize the utilization efficiency of parking resources, reduce vacant parking spaces and localized congestion. For example, guiding vehicles to remote or fixed areas can balance traffic density. As a result, the parking lot not only improves space management efficiency but can also regulate traffic flow structure through incentive strategies, thereby improving operational efficiency.

[0038] Reference Appendix Figure 2 A parking lot vehicle vacancy period assessment system according to one embodiment of the present invention includes: The acquisition module 10 is used to acquire parking image information of the parking space based on the camera device preset in the parking space at the parking terminal; The judgment module 20 is used to determine whether the parking image information detects that the parking space is covered by a shadow; The execution module 30 is configured to, if so, extract the corresponding environmental disturbance amplitude sequence from the parking image information, construct trend data of the environmental disturbance amplitude sequence, identify the fine-tuning behavior of the vehicle within a preset range of the parking space based on the trend data, and obtain the vehicle's movement trajectory. The environmental disturbance amplitude sequence specifically includes light field disturbance gradient, background texture occlusion rate, and particle motion change rate. The fine-tuning behavior specifically includes the vehicle's pixel-level distortion trajectory caused by steering wheel swing, multiple repetitions of the black and white dot tracking path, and alternating rhythm of movement and stillness. The second judgment module 40 is used to determine whether the movement trajectory matches the parking path preset by the parking terminal, wherein the parking path specifically includes parking action, trajectory overlap rate and final entry angle. The second execution module 50 is used to capture parking features in the parking image information within a preset time period if a match is found, dynamically establish a parking order for the vehicle based on the parking features, and generate parking content for the vehicle at the parking lot terminal based on the parking order. The parking features specifically include shadow boundaries, headlight illumination areas, and ground reflection patterns, and the parking content specifically includes vehicle information, parking duration, and parking space number.

[0039] In this embodiment, the acquisition module 10 acquires parking image information of a parking space based on a pre-installed and calibrated camera device in the parking lot terminal. Then, the judgment module 20 determines whether the parking image information detects that the parking space is covered by a shadow, and executes the corresponding steps accordingly. For example, when the system determines that the parking image information of the parking space is not covered by a shadow, the system considers the lighting conditions in the parking space area to be stable, and the ground texture, parking space markings, and vehicle outlines to be highly visible. The system can obtain vehicle shape and whether it occupies a space through conventional visual recognition. The system directly performs vehicle outline extraction, vehicle body boundary recognition, and license plate / vehicle volume analysis on the parking image information, and combines the parking space markings to perform geometric judgment on the vehicle's positional relationship, thereby quickly determining whether the vehicle has actually entered the parking space. Simultaneously, a fast path execution mode is adopted to complete vehicle detection and parking space occupancy judgment more efficiently. Once the vehicle has entered a steady-state position, it directly enters the behavior confirmation stage and the parking order establishment stage, thereby avoiding unnecessary process calculations and improving the system response speed and judgment accuracy. For example, when the system determines that the parking image information of the parking space is affected by shadows, the execution module 30 will consider that the lighting conditions of the parking space area are unstable and cannot be determined by conventional visual recognition. The system will extract the corresponding environmental disturbance amplitude sequence from the parking image information. The environmental disturbance amplitude sequence specifically includes the light field disturbance gradient, the background texture occlusion rate, and the particle motion change rate. The system constructs trend data of these environmental disturbance amplitude sequences. Based on different trend data, the system identifies the vehicle's fine-tuning behavior within the pre-set range of the parking space. The fine-tuning behavior specifically includes the vehicle's pixel-level distortion trajectory caused by steering wheel swing, the repeated black and white dot tracking path, and the rhythm of alternating movement and stillness, thereby obtaining the vehicle's movement trajectory.The system, by not immediately recognizing vehicles upon detecting shadow interference, first extracts the light field perturbation gradient, background texture occlusion rate, and particle motion change rate, and constructs environmental perturbation trend data. This allows it to indirectly determine parking space status even under unstable lighting or partial occlusion conditions, thus avoiding visual misjudgments caused by phenomena such as light projection, vehicle shadows, or occlusion from passing vehicles, improving the robustness and stability of parking space status judgment. Furthermore, by mapping the vehicle's motion trajectory under shadow interference to pixel-level distorted trajectories, black-and-white dot tracking paths, and dynamic / static rhythm changes, the system can identify the motion patterns of real vehicles making fine-tuning adjustments during parking from the image's underlying structure and pixel perturbations. Even when the vehicle's outline is covered by shadows and the body texture is blurred, the system can still identify changes in the vehicle's posture through micro-displacement data, achieving accurate judgment of the vehicle's presence, rather than relying on brightness features or outline clarity. Furthermore, by employing a comprehensive analysis method combining "movement trajectory + environmental trends + fine-grained pixel twisting" under shadow interference, the system can distinguish between short pauses and actual parking behaviors, ensuring that parking behavior is only confirmed when a parking path with continuity, purpose, and convergence is detected. Then, the second judgment module 40 determines whether the vehicle's movement trajectory matches the parking path pre-set in the parking lot terminal. The parking path specifically includes parking actions, trajectory overlap rate, and final parking angle. The system executes the corresponding steps. For example, when the system determines that the vehicle's movement trajectory cannot match the parking path pre-set in the parking lot terminal, the system considers that the vehicle's current behavior does not have the convergence and goal-orientation characteristics of parking behavior. This means that the vehicle is only temporarily stopping, turning around, slowly detouring, adjusting its route, or briefly observing, and is not actually performing a parking operation. The system will continue to monitor changes in the vehicle's trajectory but mark it as "non-parking path behavior," while maintaining the parking space status as "not officially occupied." The system will still show the parking space as available for other vehicles or system scheduling, thereby avoiding misjudging parking space occupancy due to short stops. Furthermore, the system combines vehicle speed, rate of change of movement direction, and stopping duration... The system further assesses whether the behavior could be transformed into a parking trend by considering factors such as the duration of the parking session and the presence of repeated attempts to enter the parking space. For example, when the system determines that the vehicle's movement trajectory matches a pre-set parking path in the parking terminal, the second execution module 50 assumes that the vehicle is moving into a parking space. The system will capture parking features in the parking image information within a pre-set time period. These parking features include shadow boundaries, headlight illumination areas, and ground reflection patterns. Based on these parking features, the system dynamically creates a parking order for the vehicle. Based on this parking order, the system generates parking content for the vehicle in the parking terminal. The parking content includes vehicle information, parking duration, and parking space number.The system effectively distinguishes between "real parking" and "temporary stops, brief stops, or detours" by using parking trajectory, entry angle, and parking action characteristics as criteria. This avoids mis-marking of parking spaces due to short-term vehicle stops, significantly improving the accuracy of parking space usage status recognition. Simultaneously, the system captures visual features that stably characterize vehicle parking status within a limited time period, such as shadow boundaries, headlight illumination areas, and ground reflection patterns. It dynamically generates parking orders based on these long-term, time-converging parking features. This order generation method, based on "continuous parking evidence," does not rely on single-frame vehicle status but utilizes the stable features of continuous images, significantly improving the reliability of parking status confirmation and reducing misjudgments caused by changes in lighting, shadow jitter, or vehicle minor adjustments. Furthermore, after generating a parking order, the system automatically creates complete parking information, including vehicle information, parking duration, and parking space number, enabling the parking lot terminal to monitor the actual occupancy of each parking space, the attributes of parked vehicles, and the status of available resources in real time.

[0040] In this embodiment, it also includes: The identification module is used to identify the blind spot area of ​​the parking space that is blocked by other vehicles, based on the preset camera angle of the camera device; The third judgment module is used to determine whether the area of ​​the camera blind spot exceeds the parking evaluation threshold of the camera device for the parking space; The third execution module is used to, if so, collect the occupancy ratio of the camera blind spot area to the parking space according to the preset reference structure line of the parking space, construct a virtual parking image when the virtual vehicle is parked in the parking space according to the occupancy ratio, and generate the occupancy disturbance signal of the parking lot terminal to the parking space through the virtual parking image. The reference structure line specifically includes the parking space boundary line, the parking space depth line and the vehicle front line.

[0041] In this embodiment, the system identifies the blind spot area of ​​a parking space obscured by other parked vehicles based on the pre-set camera angle of the camera device. The system then determines whether this blind spot area exceeds the pre-set parking evaluation threshold of the camera device for the parking space, and executes the corresponding steps accordingly. For example, if the system determines that the blind spot area of ​​the parking space obscured by other parked vehicles does not exceed the pre-set parking evaluation threshold, the system considers that the obstruction has not seriously affected the key identification area of ​​the parking space. That is, the main identification area used to determine whether a vehicle is parked in the parking space is still within the effective monitoring range. The system will use normal weights to identify the vehicle status in the still visible area. For example, when the rear area of ​​the parking space is obscured, the system prioritizes using the front of the parking space and the left and right marking directions for vehicle identification, while maintaining normal image acquisition and movement trajectory analysis of the parking space, and continuing to monitor identification indicators such as the movement trajectory of vehicles entering the parking space, changes in vehicle body pixels, and the rate of background texture occlusion. The system ensures that the parking space identification process is not interrupted due to minor occlusions, and marks the parking space as "identifiable" rather than "identifiable but obstructed" in the system's state logic. This allows the parking space to continue participating in parking guidance, route planning, and parking space availability display, preventing minor occlusions from causing the parking space to be incorrectly hidden or marked as unavailable in the system. For example, if the system determines that the area of ​​the camera blind spot obstructing the parking space exceeds the pre-set parking evaluation threshold of the camera equipment, the system will consider the obstruction to affect the key identification area of ​​the parking space. Based on the pre-set baseline structure lines for the parking space (including the parking space boundary line, parking space depth line, and vehicle front line), the system will collect the occupancy ratio of these camera blind spots. Based on different occupancy ratios, a virtual parking image of a virtual vehicle parked in the parking space will be constructed. This virtual parking image will then generate an occupancy disturbance signal for the parking space from the parking terminal.The system determines whether the occlusion area exceeds an evaluation threshold and introduces baseline structural lines (including parking space boundary lines, parking space depth lines, and vehicle front lines) for alternative recognition. It no longer relies on a complete image within the actual visual range but can infer based on structured coordinate relationships. In cases of severe occlusion, it simulates the vehicle's actual parking state using virtual parking images, avoiding the situation where "the camera's view is incomplete → the system cannot judge → the parking space status becomes invalid." This significantly improves the system's recognition stability and fault tolerance in complex occlusion environments. Furthermore, by calculating the proportion of blind spots and constructing virtual parking images, it can still generate images even when the actual footage is unavailable. This generates complete semantic images of parking spaces, ensuring that parking spaces remain "monitorable, analyzable, and assessable" at the system's internal data level. This means that even if a parking space is severely obstructed by another vehicle, the system doesn't need to hide or mark it as invalid in the guidance interface; instead, it can continue to be used for real-time scheduling, fundamentally improving the utilization efficiency of parking resources. Furthermore, the generated obstruction disturbance signals are used not only for current assessments but also for future risk predictions. For example, if a parking space is severely obstructed for an extended period, the system can mark it as a "high-risk obstruction space," thereby reducing its recommendation weight in subsequent parking route planning or issuing adjustment suggestions to staff.

[0042] In this embodiment, the execution module further includes: The identification unit is used to identify the vehicle posture vector corresponding to the parking image information based on the fine-tuning sensitive area preset by the parking terminal for the vehicle. The fine-tuning sensitive area specifically includes the front and rear bumpers, the wheel rims on both sides, and the side walls of the vehicle body. The judgment unit is used to determine whether the vehicle attitude vector exhibits a preset continuous small change; The execution unit is configured to, if so, extract the deformation parameters of the outer contour of the vehicle's tires in consecutive frames from the parking image information, calculate the micro-change value of the tire rotation angle based on the deformation parameters, obtain the number of moving pixels of the vehicle based on the vehicle body boundary in the parking image information, and generate the micro-perturbation action of the vehicle through the number of moving pixels, wherein the micro-perturbation action specifically includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.

[0043] In this embodiment, the system identifies the vehicle's attitude vector corresponding to the parking image information based on a pre-defined fine-tuning sensitive area for the vehicle at the parking terminal. This sensitive area specifically includes the front and rear bumpers, wheel rims, and sidewalls. The system then determines whether these attitude vectors exhibit a pre-defined continuous small change to execute corresponding steps. For example, if the system determines that the vehicle's attitude vector in the parking image information does not exhibit a pre-defined continuous small change, the system considers that the vehicle has not performed any obvious fine-tuning action, meaning the vehicle is in a stable parking state or has not actually entered the parking action phase. The system then directly enters the vehicle stability judgment process, comparing the current attitude vector and parking position with the parking space baseline to confirm whether the vehicle is fully in place. Simultaneously, based on the stable vehicle attitude, the system performs normal parking recording and order generation operations for the parking space. The system collects parking features and generates parking orders and parking content (vehicle information, parking duration, parking space number), and maintains continuous monitoring of vehicles. If a vehicle subsequently exhibits new posture changes, its parking or fine-tuning status is reassessed to ensure the real-time and accuracy of parking behavior judgment. For example, when the system determines that the vehicle posture vector corresponding to a certain vehicle in the parking image information shows a pre-set continuous small change, the system will consider that the vehicle has made a fine-tuning action in the parking space. The system will extract the deformation parameters of the outer contour of the vehicle's tires in consecutive frames from the parking image information, calculate the micro-change value of the tire rotation angle based on different deformation parameters, obtain the number of moving pixels of the vehicle based on the vehicle body boundary in the parking image information, and generate the vehicle's micro-perturbation action through these moving pixels. The micro-perturbation action specifically includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.By extracting the deformation parameters of the vehicle's outer tire contour in consecutive frames and combining the slight changes in tire angle with the pixel movement of the vehicle body boundary, the system can accurately identify the vehicle's subtle adjustments in a parking space, such as slow forward and backward movement, slight straightening of the front of the vehicle, or slight lateral movement. Compared with traditional methods that rely solely on the overall vehicle body contour or single-frame brightness information, this method can capture subtle changes in vehicle posture, significantly reducing misjudgments of parking status caused by undetected subtle adjustments. Furthermore, even in complex environments such as uneven lighting, shadow coverage, or occlusion by adjacent vehicles, subtle adjustments in vehicle posture can still be captured by the system through tire contour deformation and pixel-level movement. This method makes the system... Even under visual interference, the system can still determine whether a vehicle is completing a parking maneuver, thus improving the stability and reliability of parking space occupancy assessment. This avoids misclassifying short-term or temporary parking as actual parking. Furthermore, by recognizing subtle vehicle movements, the system can more accurately determine the vehicle's final parking status and dynamically adjust parking orders accordingly. Through this subtle movement information, the parking lot terminal can promptly record the time, location, and posture changes of a vehicle upon completion of parking. This enhances the accuracy and traceability of the parking management system in areas such as billing, parking guidance, inspection, and abnormal behavior monitoring, thereby improving the overall intelligent operation level of the parking lot.

[0044] In this embodiment, the second execution module further includes: The detection unit is used to detect the departure time of the vehicle in the parking space based on the vehicle characteristics pre-collected by the parking terminal. The vehicle characteristics specifically include vehicle model, license plate data and vehicle appearance. The second judgment unit is used to determine whether the departure period can be settled within a preset time period; The second execution unit is used to identify the cumulative number of parking spaces occupied by the vehicle in the parking lot if no, generate the pending settlement information of the vehicle in the parking lot based on the cumulative number of parking spaces, and dynamically reset the billing period of the parking space based on the pending settlement information.

[0045] In this embodiment, the system detects the departure time of a vehicle in a parking space based on vehicle characteristics pre-collected by the parking lot terminal. These characteristics include vehicle model, license plate data, and vehicle appearance. The system then determines whether the departure time can be settled within a pre-set duration and executes the corresponding steps accordingly. For example, if the system determines that a vehicle can be settled within a pre-set duration after leaving the parking space, it considers the vehicle's departure behavior to comply with the parking lot's billing and settlement rules. That is, the parking time is within the system's preset settlement period, the vehicle has not exceeded the time limit, or there is no abnormal departure behavior. The system then initiates the parking settlement process based on the vehicle's parking time in the parking space, the pre-collected vehicle characteristics, and the pre-collected vehicle characteristics. The system calculates the amount due based on fee rules and discount policies, and records information such as vehicle departure time, parking duration, settlement amount, and parking space number to the parking management terminal, generating a complete departure order. This order is linked to the vehicle's terminal to ensure subsequent data traceability. After a vehicle completes settlement upon departure, the system immediately updates the corresponding parking space status to "available" for subsequent vehicles to use. It also sends parking space availability information to the parking guidance system or management terminal, achieving dynamic management and efficient utilization of parking resources. For example, if the system determines that a vehicle has left a parking space but has not settled its departure within the pre-set time, the system will consider that the vehicle lingered in the parking lot after leaving the parking space, possibly due to another vehicle being used. When a vehicle parks in a designated spot, the system identifies the vehicle's cumulative parking space count within the parking lot. Based on this count, it generates pending settlement information for the vehicle and dynamically resets the billing period for each parking space based on the specific pending settlement information. The system can also identify continuous lingering behavior in the parking lot by determining if a vehicle fails to complete settlement within a preset time after leaving a parking space. This includes situations such as changing parking spaces or short stops. Based on the cumulative parking space count, it generates pending settlement information to ensure that every parking activity is recorded and billed. This effectively prevents missed parking fees due to short-term vehicle movement or changing spaces, improving the completeness and accuracy of parking lot settlements. Furthermore, it dynamically resets the billing period for each parking space based on different pending settlement information. By setting a billing period for parking spaces, the continuous parking behavior of the same vehicle in the parking lot can be reasonably split or combined for billing, ensuring that the billing logic is consistent with the actual parking behavior. This method can handle complex situations such as vehicles changing parking spaces multiple times or making short stops, improving the billing flexibility and intelligent management capabilities of the parking lot system. Furthermore, by identifying the cumulative number of parking spaces a vehicle has parked in the parking lot and generating dynamic pending settlement information, the system can update the usage status and availability of each parking space in real time, avoiding resource waste caused by long-term unsettled parking spaces. This mechanism not only ensures the accuracy of parking space status information, but also provides reliable data for guiding parking, reserving parking spaces, and parking space scheduling, improving the overall operational efficiency and user experience of the parking lot.

[0046] In this embodiment, the determination module further includes: The second identification unit is used to identify the shadow type of the shadow coverage area based on the shadow coverage area pre-detected by the parking terminal for the parking space. Specifically, the shadow type includes static structural shadows, dynamic lighting shadows, and projection changes caused by light source flicker. The third judgment unit is used to determine whether the proportion of the shadow-covered area exceeds the preset proportion of the parking space; The third execution unit is configured to, if so, construct the shadow area expansion rate when the vehicle is parked based on the shadow type, dynamically predict the shadow movement trend of the parking space based on the shadow area expansion rate, and mark the occlusion area of ​​the vehicle when it is parked in the parking space based on the shadow movement trend.

[0047] In this embodiment, the system identifies the shadow types of parking spaces based on the shadow coverage areas pre-detected by the parking terminal. Specific shadow types include static structural shadows, dynamic lighting shadows, and projection changes caused by light source flicker. The system then determines whether the proportion of these shadow coverage areas exceeds a pre-set upper limit for the parking space, and executes corresponding steps accordingly. For example, if the system determines that the proportion of the shadow coverage area of ​​a parking space does not exceed the pre-set upper limit, the system considers the lighting conditions of that parking space to be within an acceptable range, and the shadows have a minimal impact on vehicle recognition and parking status assessment. The system will then continue to use the standard visual recognition process based on vehicle outlines, parking space markings, shadow boundaries, etc. The system determines whether a parking space is occupied by capturing vehicle movement trajectories, posture changes, and parking characteristics in consecutive frames. It monitors vehicle entry and fine-tuning behavior, and uses shadow information to remove trends in pixel brightness changes, ensuring accuracy. Since the shadow does not substantially affect key recognition areas, the system marks the parking space as "identifiable and available," allowing it to participate in parking guidance, route planning, and management scheduling, ensuring efficient use of parking resources. For example, if the system determines that the proportion of a parking space's shadow coverage exceeds the pre-set limit, it considers the lighting conditions of that parking space excessively obscured by shadows, affecting the camera's judgment of the parking status. The system will then adjust its approach based on different shadow types. The system constructs the shadow area expansion rate when the vehicle is parked. Based on these expansion rates, it dynamically predicts the shadow movement trend of the parking space and marks the occluded area of ​​the vehicle when parked. The system can identify severely affected lighting conditions in the parking space when the shadow coverage ratio exceeds a preset threshold. For different types of shadows, the system constructs the shadow area expansion rate when the vehicle is parked. This expansion rate quantifies the degree of occlusion of the vehicle outline and key areas of the parking space, providing a quantifiable reference for subsequent vehicle recognition. This reduces the risk of misjudgment caused by lighting interference or shadow coverage. Furthermore, by constructing shadow movement trends using the shadow area expansion rate, the system can predict potential occlusion changes during vehicle parking. The system tracks the movement and expansion of shadow boundaries over time. Based on this trend, the system can mark areas that may be obscured when a vehicle is parked in real time, providing compensation information for fine-tuning action analysis, tire posture recognition, and vehicle body contour detection. This ensures the stability and reliability of vehicle parking status judgment even under complex lighting conditions. Furthermore, through quantitative and dynamic analysis of shadow coverage areas and occupancy trends, the system can accurately determine the vehicle's parking and fine-tuning status in parking spaces with complex lighting conditions, ensuring real-time updates of parking space occupancy status. The generated occupancy area information can be used for parking guidance, route planning, and abnormal parking monitoring, enabling the parking management system to operate efficiently and intelligently in complex environments, improving the utilization efficiency of parking resources and user experience.

[0048] In this embodiment, the second determination module further includes: The acquisition unit is used to acquire continuous frame images of the vehicle entering the parking space based on the preset parking space geometry. The parking space geometry specifically includes a front parking area, a vehicle body alignment area, and a rear calibration area. The fourth judgment unit is used to determine whether the continuous frame image is acquired in segments; The fourth execution unit is configured to, if so, identify the vehicle's contact area with the parking space, obtain the vehicle's spatial position relative to adjacent parking spaces based on the contact area, dynamically generate the optimal parking angle for the vehicle in the parking space based on the spatial position, and send the optimal parking angle to the vehicle through the parking lot terminal.

[0049] In this embodiment, the system acquires continuous frame images of a vehicle entering a parking space based on a pre-set parking space geometry, specifically including a front entry area, a body alignment area, and a rear calibration area. The system then determines whether these continuous frame images were acquired in segments to execute corresponding steps. For example, if the system determines that the continuous frame images of a vehicle entering a parking space were not acquired in segments, the system assumes that the vehicle did not exhibit any obvious staged fine-tuning movements during the entire entry process. The vehicle is considered to have been parked directly in the parking space without any directional adjustments or forward / backward movement. The system will directly match the image with the parking space geometry without needing to analyze the fine-tuning trajectory or tire posture changes, quickly confirming the vehicle's position. The vehicle has been correctly positioned, and its parking status has been recorded. Upon completion of the parking process, the system directly generates a parking order, including vehicle information, entry time, parking space number, and parking status, and marks the parking space as "occupied." This allows the parking management system to update parking space status in real time, ensuring smooth operation of parking guidance, route planning, and billing processes. For example, if the system determines that consecutive frame images of a vehicle entering a parking space are segmented, it assumes the vehicle is adjusting its parking space to avoid improper parking. The system identifies the vehicle's contact area with the parking space and, based on different contact areas, obtains the vehicle's spatial position relative to adjacent parking spaces. The system dynamically generates the optimal parking angle for the vehicle within the parking space and sends this angle to the vehicle via a pre-established communication channel between the vehicle and the parking lot terminal. By analyzing segmented, continuous frame images of the vehicle within the parking space, the system can determine if the vehicle has made minor adjustments during the parking process to adapt to the available space. Based on the relationship between the vehicle and the contact area of ​​the parking space, the system calculates the vehicle's spatial position relative to adjacent parking spaces using camera equipment, dynamically generating the optimal parking angle. This guides the vehicle to park precisely within a limited space, reducing the risk of scratches, misalignment, or occupying multiple parking spaces due to improper parking angles. Simultaneously, the system sends the optimal parking angle to the vehicle via a pre-established communication channel, ensuring the vehicle... The system allows vehicles to adjust their direction and parking strategy in real time. This information feedback mechanism provides intelligent assistance during the actual parking process, not only improving the convenience of driver operation but also enhancing the parking lot's real-time control capabilities in complex environments (such as adjacent vehicles, narrow parking spaces, or obstructions from pillars). This makes parking behavior more standardized and efficient. Furthermore, by dynamically calculating the spatial position of the vehicle relative to adjacent parking spaces and generating the optimal parking angle, the system ensures that each vehicle is parked in the center of the parking space at a suitable angle, maximizing parking space utilization efficiency and reducing waste of adjacent spaces caused by vehicle misalignment. By precisely guiding vehicles to fine-tune their parking, the system can reduce the risk of vehicle collisions and the incidence of parking accidents, thereby improving the overall management level and operational safety of the parking lot.

[0050] In this embodiment, it also includes: The acquisition module is used to acquire an empty parking space area in the parking lot based on the real-time location of the vehicle pre-identified by the parking lot terminal, wherein the empty parking space area is specifically an vacant parking space that has been settled. The fourth judgment module is used to determine whether the vehicle is parked in the empty space area; The fourth execution module is used to, if so, construct parking discount information for the vehicle based on the parking status of the vehicle, and dynamically update the parking fee of the vehicle based on the parking discount information.

[0051] In this embodiment, the system obtains vacant parking spaces within the parking lot based on the real-time location of the vehicle pre-identified by the parking lot terminal. Specifically, vacant parking spaces are those that are empty and have been settled. The system then determines whether a vehicle is parked within a vacant parking space and executes the corresponding steps accordingly. For example, if the system determines that a vehicle has not parked in the vacant parking space assigned to it by the parking lot terminal after entering the lot, the system considers that the vehicle has not parked according to the parking lot terminal's scheduling instructions. This may indicate arbitrary parking, illegal parking, disorderly occupation of spaces, or temporary parking. The system will then re-mark the vehicle's current actual parking location based on the vehicle's real-time location trajectory and update the status of the corresponding parking space. The status is updated to "abnormal occupancy" to prevent the parking space from being misled into being "vacant and available." Simultaneously, a notification message is automatically generated indicating off-center parking or parking outside designated areas. This message is sent to the parking lot's reminder terminal via license plate association or pushed to the driver through a pre-set communication channel, reminding the driver that their vehicle is not parked in the designated space. Furthermore, the parking resources are reassessed based on the vehicle's actual parking location. The vacant area originally allocated to that vehicle is re-released as available, and guidance information for other entering vehicles is dynamically updated to ensure the rational use of parking resources and prevent confusion caused by the vehicle leading to vacancy or occupancy of existing spaces. For example, when the system detects a vehicle parking outside designated areas after entering the parking lot... When a vehicle is parked in an empty space assigned by the parking terminal, the system considers it to be parked in accordance with the terminal's dispatch instructions. Based on the vehicle's parking status, the system constructs parking discount information and dynamically updates the parking fee accordingly. By recognizing that a vehicle is indeed parked in a system-assigned empty space, the system rewards the driver for correctly following the parking terminal's dispatch instructions. By incentivizing standardized parking behavior through parking discount information, this mechanism effectively guides vehicles to follow parking guidelines, thereby reducing random parking, occupying emergency areas, or parking out of bounds, thus creating an orderly parking environment throughout the parking lot. Parking fees are dynamically updated based on parking status. A reward mechanism directly links parking costs to the degree of compliance with parking rules. This dynamic billing model not only improves the parking experience but also enhances the interaction between drivers and the parking system, making users more willing to cooperate and follow guidance, thus increasing user satisfaction and the self-discipline of parking behavior. Furthermore, by incentivizing vehicles to prioritize parking in system-assigned empty spaces, the system can maximize the utilization efficiency of parking resources, reduce vacant parking spaces and localized congestion. For example, guiding vehicles to remote or fixed areas can balance traffic density. As a result, the parking lot not only improves space management efficiency but can also regulate traffic flow structure through incentive strategies, thereby improving operational efficiency.

[0052] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the cyclical availability of parking spaces, characterized in that, Includes the following steps: Based on the parking terminal's pre-set camera equipment for parking spaces, the parking image information of the parking spaces is collected; Determine whether the parking image information detects that the parking space is covered by a shadow; If so, the corresponding environmental disturbance amplitude sequence is extracted from the parking image information, and trend data of the environmental disturbance amplitude sequence is constructed. Based on the trend data, the fine-tuning behavior of the vehicle within the preset range of the parking space is identified, and the movement trajectory of the vehicle is obtained. The environmental disturbance amplitude sequence specifically includes light field disturbance gradient, background texture occlusion rate and particle motion change rate. The fine-tuning behavior specifically includes the vehicle pixel-level distortion trajectory caused by steering wheel swing, the repeated black and white dot tracking path, and the rhythm of alternating movement and stillness. Determine whether the movement trajectory matches the parking path preset by the parking terminal, wherein the parking path specifically includes parking action, trajectory overlap rate and final entry angle; If a match is found, parking features in the parking image information are captured within a preset time period. Based on the parking features, a parking order for the vehicle is dynamically created. Based on the parking order, parking content for the vehicle is generated at the parking lot terminal. The parking features specifically include shadow boundaries, headlight illumination areas, and ground reflection patterns. The parking content specifically includes vehicle information, parking duration, and parking space number.

2. The method for evaluating the cyclical availability of parking spaces in a parking lot according to claim 1, characterized in that, Before the step of extracting the corresponding environmental disturbance amplitude sequence from the parking image information and constructing the trend data of the environmental disturbance amplitude sequence, the method further includes: Based on the preset camera angle of the camera device, identify the blind spot area of ​​the parking space that is blocked by other parked vehicles; Determine whether the area of ​​the camera blind spot exceeds the parking assessment threshold of the camera device for the parking space; If so, based on the preset baseline structure line of the parking space, the proportion of the camera blind spot area occupying the parking space is collected. Based on the occupancy proportion, a virtual parking image of a virtual vehicle parked in the parking space is constructed. Through the virtual parking image, the occupancy disturbance signal of the parking lot terminal on the parking space is generated. The baseline structure line specifically includes the parking space boundary line, the parking space depth line, and the vehicle front line.

3. The method for evaluating the cyclical availability of parking spaces in a parking lot according to claim 1, characterized in that, The step of identifying the vehicle's fine-tuning behavior within the preset range of the parking space also includes: Based on the parking terminal's preset fine-tuning sensitive area for the vehicle, the vehicle posture vector corresponding to the parking image information is identified, wherein the fine-tuning sensitive area specifically includes the front and rear bumpers, the two side wheel rims and the vehicle body sidewalls; Determine whether the vehicle attitude vector exhibits a preset, continuous, small change; If so, the deformation parameters of the outer contour of the vehicle's tires in consecutive frames are extracted from the parking image information. Based on the deformation parameters, the micro-change value of the tire rotation angle is calculated. Based on the vehicle body boundary in the parking image information, the number of moving pixels of the vehicle is obtained. Based on the number of moving pixels, the micro-perturbation action of the vehicle is generated. Specifically, the micro-perturbation action includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.

4. The method for evaluating the cyclical availability of parking spaces in a parking lot according to claim 1, characterized in that, The step of generating the vehicle's parking information on the parking terminal based on the parking order further includes: Based on the vehicle characteristics pre-collected by the parking terminal for the parking space, the departure time of the vehicle in the parking space is detected. The vehicle characteristics specifically include vehicle model, license plate data and vehicle appearance. Determine whether the departure period can be settled within a preset time limit; If not, the system identifies the cumulative number of parking spaces the vehicle has occupied in the parking lot, generates pending settlement information for the vehicle in the parking lot based on the cumulative number of parking spaces, and dynamically resets the billing period for the parking spaces based on the pending settlement information.

5. The method for evaluating the cyclical availability of parking spaces in a parking lot according to claim 1, characterized in that, The step of determining whether the parking image information detects that the parking space is covered by a shadow also includes: Based on the shadow coverage area pre-detected by the parking terminal for the parking space, the shadow type of the shadow coverage area is identified, wherein the shadow type specifically includes static structural shadow, dynamic lighting shadow and projection change caused by light source flicker; Determine whether the proportion of the shaded area exceeds the preset proportion of the parking space; If so, then based on the shadow type, construct the shadow area expansion rate when the vehicle is parked, dynamically predict the shadow movement trend of the parking space based on the shadow area expansion rate, and mark the occlusion area of ​​the vehicle when it is parked in the parking space based on the shadow movement trend.

6. The method for evaluating the cyclical availability of parking spaces in a parking lot according to claim 1, characterized in that, The step of determining whether the movement trajectory matches the parking path preset by the parking terminal further includes: Based on the preset parking space geometry, continuous frame images of the vehicle entering the parking space are acquired, wherein the parking space geometry specifically includes a front parking area, a vehicle body alignment area, and a rear calibration area. Determine whether the continuous frame images were acquired in segments; If so, the system identifies the vehicle's contact area with the parking space, obtains the vehicle's spatial position relative to adjacent parking spaces based on the contact area, dynamically generates the optimal parking angle for the vehicle in the parking space based on the spatial position, and sends the optimal parking angle to the vehicle through the parking lot terminal.

7. The method for evaluating the cyclical availability of parking spaces in a parking lot according to claim 1, characterized in that, Before the step of acquiring parking image information of the parking space using a camera device preset in the parking space based on the parking terminal, the method further includes: Based on the real-time location of the vehicle pre-identified by the parking terminal, an empty parking space area is obtained in the parking lot, wherein the empty parking space area is specifically an vacant parking space that has been settled. Determine whether the vehicle is parked in the empty space area; If so, then based on the parking status of the vehicle, parking discount information for the vehicle is constructed, and the parking fee for the vehicle is dynamically updated based on the parking discount information.

8. A parking lot vehicle vacancy period assessment system, characterized in that, include: The acquisition module is used to acquire parking image information of the parking space based on the camera device preset in the parking space at the parking terminal; The judgment module is used to determine whether the parking image information detects that the parking space is covered by a shadow; The execution module is configured to, if so, extract the corresponding environmental disturbance amplitude sequence from the parking image information, construct trend data of the environmental disturbance amplitude sequence, identify the fine-tuning behavior of the vehicle within a preset range of the parking space based on the trend data, and obtain the vehicle's movement trajectory. Specifically, the environmental disturbance amplitude sequence includes light field disturbance gradient, background texture occlusion rate, and particle motion change rate. The fine-tuning behavior specifically includes the vehicle's pixel-level distortion trajectory caused by steering wheel swing, multiple repetitions of the black and white dot tracking path, and alternating rhythms of movement and stillness. The second judgment module is used to determine whether the movement trajectory matches the parking path preset by the parking terminal, wherein the parking path specifically includes parking action, trajectory overlap rate and final entry angle. The second execution module is used to capture parking features in the parking image information within a preset time period if a match is found, dynamically establish a parking order for the vehicle based on the parking features, and generate parking content for the vehicle at the parking lot terminal based on the parking order. The parking features specifically include shadow boundaries, headlight illumination areas, and ground reflection patterns, and the parking content specifically includes vehicle information, parking duration, and parking space number.

9. The parking lot vehicle vacancy period assessment system according to claim 8, characterized in that, Also includes: The identification module is used to identify the blind spot area of ​​the parking space that is blocked by other vehicles, based on the preset camera angle of the camera device; The third judgment module is used to determine whether the area of ​​the camera blind spot exceeds the parking evaluation threshold of the camera device for the parking space; The third execution module is used to, if so, collect the occupancy ratio of the camera blind spot area to the parking space according to the preset reference structure line of the parking space, construct a virtual parking image when the virtual vehicle is parked in the parking space according to the occupancy ratio, and generate the occupancy disturbance signal of the parking lot terminal to the parking space through the virtual parking image. The reference structure line specifically includes the parking space boundary line, the parking space depth line and the vehicle front line.

10. The parking lot vehicle vacancy period assessment system according to claim 8, characterized in that, The execution module further includes: The identification unit is used to identify the vehicle posture vector corresponding to the parking image information based on the fine-tuning sensitive area preset by the parking terminal for the vehicle. The fine-tuning sensitive area specifically includes the front and rear bumpers, the wheel rims on both sides, and the side walls of the vehicle body. The judgment unit is used to determine whether the vehicle attitude vector exhibits a preset continuous small change; The execution unit is configured to, if so, extract the deformation parameters of the outer contour of the vehicle's tires in consecutive frames from the parking image information, calculate the micro-change value of the tire rotation angle based on the deformation parameters, obtain the number of moving pixels of the vehicle based on the vehicle body boundary in the parking image information, and generate the micro-perturbation action of the vehicle through the number of moving pixels, wherein the micro-perturbation action specifically includes slow forward and backward movement, slight straightening of the front of the vehicle, and slight lateral movement.

Images