Carpet atmosphere lamp and automobile with same

Abstract

The invention provides a carpet atmosphere lamp and an automobile with the carpet atmosphere lamp, and the carpet atmosphere lamp comprises atmosphere lamps which are installed at the two sides of an automobile chassis and are used for generating a virtual carpet; the detection module is used for acquiring getting-on and getting-off data of a user and generating trajectory prediction data according to the getting-on and getting-off data; the instruction generation module is used for receiving the trajectory prediction data and performing matching in a preset control instruction library according to the trajectory prediction data to obtain an atmosphere lamp control instruction; and the atmosphere lamp control module is used for controlling the atmosphere lamp to generate a virtual carpet according to the atmosphere lamp control instruction.

Description

Technical Field

[0001] The present application relates to the technical field of lamp lighting control, and in particular to a carpet atmosphere lamp and a car having the carpet atmosphere lamp. Background Art

[0002] Currently, existing carpet ambient lighting systems typically only offer basic on / off control and limited color changes, lacking deep integration with the vehicle's overall system. This design approach fails to fully leverage the rich sensor networks and powerful computing capabilities of smart car platforms, resulting in a significant mismatch between functionality and platform capabilities. Smart cars are equipped with a variety of sensors, such as speed sensors, accelerometers, temperature sensors, light sensors, cameras, and biometric sensors. These devices collect real-time data on the vehicle's interior and exterior environments, as well as user status information. Modern smart cars also possess powerful data processing and artificial intelligence computing capabilities, enabling them to analyze and process this collected information. However, current carpet ambient lighting systems fail to effectively connect with these intelligent resources, unable to automatically adjust lighting effects based on driving scenarios, passenger habits, or external environmental factors, thus missing the opportunity to create an immersive, personalized, and intelligent in-car experience. Summary of the Invention

[0003] The present application provides a carpet atmosphere light and a car with a carpet atmosphere light, which are used to intelligently control the atmosphere light to generate a virtual carpet that matches the current environment according to passenger habits and the external environment, so as to improve the user experience.

[0004] In a first aspect, an embodiment of the present application provides a carpet atmosphere lamp, comprising: Ambient lights, which are installed on both sides of the car chassis and are used to generate a virtual carpet; A detection module, configured to obtain the user's boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data; An instruction generation module is used to receive the trajectory prediction data and match it with a preset control instruction library according to the trajectory prediction data to obtain an atmosphere light control instruction; The atmosphere light control module is used to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction.

[0005] In a second aspect, an embodiment of the present application provides a car with a carpet atmosphere light, wherein the car is equipped with a carpet atmosphere as described in any one of the embodiments of the present application, and the car includes: Ambient lights, which are installed on both sides of the car chassis and are used to generate a virtual carpet; A detection module, configured to obtain the user's boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data; An instruction generation module is used to receive the trajectory prediction data and match it with a preset control instruction library according to the trajectory prediction data to obtain an atmosphere light control instruction; The atmosphere light control module is used to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction.

[0006] An embodiment of the present application provides a carpet atmosphere light, comprising: an atmosphere light, a detection module, a command generation module, and an atmosphere light control module. The atmosphere light is mounted on both sides of the vehicle chassis and is used to generate a virtual carpet. The detection module is used to obtain user boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data. The command generation module is used to receive the trajectory prediction data and match it with a preset control command library to obtain atmosphere light control commands. The atmosphere light control module is used to control the atmosphere light to generate the virtual carpet according to the atmosphere light control commands. In this carpet atmosphere light, by intelligently detecting user boarding and alighting behavior and predicting the travel trajectory, it can automatically project a virtual carpet to provide precise lighting guidance. This not only improves safety at night or in low-light environments, but also enables personalized intelligent interaction based on user behavior, automatically activating and adjusting the lighting range according to actual needs, reducing the user's operational burden and optimizing energy efficiency. This function fully utilizes the sensing and computing capabilities of smart cars, improving the overall intelligence level and brand recognition, and providing users with a more convenient, safe, and personalized car experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0008] Figure 1 A schematic block diagram of a carpet atmosphere lamp provided in an embodiment of the present application; Figure 2 A schematic block diagram of a car with carpet ambient lighting provided in an embodiment of the present application. DETAILED DESCRIPTION

[0009] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0010] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0011] It should also be understood that the terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0012] It should be further understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0013] See also Figure 1 , Figure 1 This is a schematic flow chart of a carpet atmosphere light provided by an embodiment of the present application. Figure 1 As shown, the carpet atmosphere light includes: an atmosphere light, a detection module, a command generation module and an atmosphere light control module.

[0014] Ambient lights are installed on both sides of the car chassis and are used to generate virtual carpets.

[0015] For example, the ambient lighting, the system's core output unit, is strategically located on both sides of the vehicle's chassis. Utilizing a high-brightness LED array and precision optical lenses, it projects a distinct virtual carpet effect onto the ground around the vehicle. The ambient lighting features a 360-degree motorized rotation mechanism, enabling real-time adjustment of the projection angle to follow the user's movements. A dedicated heat dissipation system ensures stable operation even at high brightness. The light source utilizes independent RGB color control technology, dynamically adjusting color temperature and hue to suit different scenarios. In welcome mode, it creates a warm, soft gradient effect, while in safety alert mode, it switches to high-contrast warning colors. The ambient lighting projection system utilizes a specialized optical structure that focuses the projected light into pre-set patterns and text. Precise control of the lens array creates a variety of virtual carpets, including fan-shaped guide light carpets, linear strips, and circular warning apertures. PWM pulse modulation technology creates precisely controlled phase differences between the multiple ambient lights, creating a flowing light pattern on the ground, enhancing the user's sense of motion. The ambient lighting system realizes a visual interaction channel between the vehicle and the user through intelligent light effect control, providing an intuitive and aesthetic human-vehicle interaction experience.

[0016] The detection module is used to obtain the user's boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data.

[0017] For example, the detection module, serving as the system's perception unit, captures comprehensive data on user boarding and alighting behaviors through a multi-sensor fusion architecture. This module achieves precise perception of user behavior through the collaborative operation of an infrared detection unit, a near-field communication unit, and a vehicle docking unit. The near-field communication unit establishes a wireless connection with the user's portable vehicle control module. Upon receiving a start-up control signal, it automatically switches to user approach detection, and upon disconnection, it switches to user departure detection, enabling proactive recognition of user intent. The infrared detection unit captures thermal imaging data, accurately locating the user's position and movement even in low-light environments. The vehicle docking unit collects vehicle hardware control information, including door lock status and door opening and closing angles. Combined with visible light images captured by the camera unit, the system generates a multimodal data set comprising infrared, visible light, and vehicle status. The detection module incorporates a pre-trained multimodal trajectory analysis model, comprising a deep learning architecture consisting of an image processing network, a text processing network, a fully connected network, and an output network. This model converts the collected heterogeneous multi-source data into standardized trajectory feature vectors and action intention feature vectors. Through latent feature extraction and fusion, it ultimately generates highly accurate user trajectory prediction data. This detection module realizes intelligent prediction of user behavior intentions through multimodal data perception and deep learning processing, providing a decision-making basis for system response.

[0018] The instruction generation module is used to receive trajectory prediction data and match it with a preset control instruction library according to the trajectory prediction data to obtain the atmosphere light control instruction.

[0019] For example, the command generation module serves as the system's decision-making hub, responsible for translating abstract trajectory prediction data into specific ambient lighting control commands. This module includes a library of preset control commands, including various contextual strategies such as welcome start commands, carpet guidance commands, safety warning commands, lateral following commands, energy-saving maintenance commands, and carpet fade-out commands. The module first extracts key parameters from the trajectory prediction data, including the user's real-time coordinates, movement velocity vector, and horizontal projection distance from the vehicle chassis. It then establishes three response zones based on the horizontal projection distance: a long-distance welcome zone of 3-5 meters, a medium-distance guidance zone of 1-3 meters, and a close-range safety zone of 0-1 meter. When the user is detected in the long-distance zone and moving toward the vehicle, the system responds to the welcome start command, activates a breathing gradient mode, and projects a welcome carpet with dynamic text. In the medium-distance zone, the system intelligently selects either linear guidance or path correction mode based on the angle between the user's movement direction and the door trajectory, providing precise directional guidance. When the user enters the close-distance zone, the system activates the safety warning mode, increasing lighting brightness and superimposing a warning light ring around the user's feet. For lateral movement, the system adjusts the light carpet's projection angle in real time, creating a tracking effect. When the user is stationary, the system automatically enters energy-saving mode, reducing brightness and shrinking the light carpet's range. As the user moves away, the system executes a phased light carpet fading sequence based on distance. This instruction generation module achieves an intelligent response mechanism for the user's entire behavior process through a precise matching strategy of distance zoning and behavior recognition.

[0020] The atmosphere light control module is used to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction.

[0021] For example, the ambient light control module, serving as the system's execution unit, converts abstract commands into actual control signals for the ambient lights. This module dynamically adjusts the projection angles of multiple ambient lights by parsing the projection mode parameters contained in the ambient light control commands. Specifically, when an angle between the user's movement direction and the door opening direction is detected, differential rotation technology is used to precisely compensate for the projection angle offset, ensuring that the virtual carpet always optimally matches the user's walking path. To meet different scenario requirements, the control module flexibly adjusts the light-emitting unit drive current based on the light intensity coefficient carried in the command. In welcome mode, a sinusoidal dimming algorithm is applied to create a natural, smooth, breathing gradient effect, while in safety warning mode, the light output is constant at maximum power to provide sufficient safety lighting. The control module also adjusts the virtual carpet's coverage based on the vehicle's door lock status and opening angle data. When the door opening angle exceeds 60 degrees, wide-area coverage mode is automatically activated, expanding the light carpet's width to 1.2 times the standard value and generating dynamic boundary markers at the bottom of the door, providing clearer guidance for boarding and alighting. In terms of dynamic lighting control, the module uses PWM pulse modulation technology to precisely adjust the phase differences between the beams of multiple ambient lights, creating a rippled, flowing light effect. The flow direction is intelligently synchronized with the user's movement speed vector, enhancing the carpet's three-dimensional and dynamic visual effects. This ambient light control module synchronizes dynamic lighting effects with precise optical projection control, achieving intelligent, adaptive display capabilities for the virtual carpet.

[0022] An embodiment of the present application provides a carpet atmosphere light, comprising: an atmosphere light, a detection module, a command generation module, and an atmosphere light control module. The atmosphere light is mounted on both sides of the vehicle chassis and is used to generate a virtual carpet. The detection module is used to obtain user boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data. The command generation module is used to receive the trajectory prediction data and match it with a preset control command library to obtain atmosphere light control commands. The atmosphere light control module is used to control the atmosphere light to generate the virtual carpet according to the atmosphere light control commands. In this carpet atmosphere light, by intelligently detecting user boarding and alighting behavior and predicting the travel trajectory, it can automatically project a virtual carpet to provide precise lighting guidance. This not only improves safety at night or in low-light environments, but also enables personalized intelligent interaction based on user behavior, automatically activating and adjusting the lighting range according to actual needs, reducing the user's operational burden and optimizing energy efficiency. This function fully utilizes the sensing and computing capabilities of smart cars, improving the overall intelligence level and brand recognition, and providing users with a more convenient, safe, and personalized car experience.

[0023] In order to more clearly introduce the technical solution of the present application, the technical solution of the present application will be introduced through specific embodiments below. It should be noted that the specific embodiments are used to expand the technical solution of the present application, but are not intended to limit the present application.

[0024] In some embodiments, the detection module is further configured to communicate with the user's portable vehicle control module. When the detection module receives a start control signal from the portable vehicle control module, it enters a user approach detection state and acquires user boarding data. When the detection module detects that the user has left the vehicle, it enters a user departure detection state and acquires user exit data.

[0025] Exemplarily, the detection module establishes a stable communication link with the user's portable vehicle control module to achieve intelligent sensing and data collection of the user's status. The portable vehicle control module is typically a smart key or mobile terminal application with integrated near-field communication (NFC) functionality. When the user brings the module within the effective communication range of the vehicle (typically 5-10 meters), the detection module's NFC unit automatically completes authentication and establishes a connection. For example, when the user approaches the vehicle's bound digital key within 3 meters, the NFC unit triggers a handshake signal via the Bluetooth Low Energy protocol and then sends a start control signal to the detection module, switching the system to user proximity detection. In this state, the detection module's multimodal sensing units begin to work together. The infrared detection unit uses a thermal imaging sensor mounted at the bottom of the vehicle door to capture the thermal radiation distribution of the user's legs, generating infrared image data with a resolution of 320×240. The vehicle docking unit reads the door lock status, seat pressure sensor values, and transmission gear position information from the vehicle's CAN bus in real time, forming a hardware control information data stream. Simultaneously, wide-angle cameras on both sides of the vehicle capture visible light images of the user's posture at a rate of 30 frames per second, using edge computing devices to identify the user's contour. Once the user enters the vehicle and closes the door, the detection module determines the user has entered by analyzing the seat pressure sensor's continuous load data (e.g., a pressure value that remains stable above 50 kg for more than 5 seconds) and the transmission's signal to switch to drive mode. During this time, the near-field communication unit remains in low-power monitoring mode. When the user stops the vehicle and opens the door, the detection module automatically triggers user exit detection based on a sudden change in the door opening sensor value (e.g., an opening angle exceeding 45 degrees) and the seat pressure release signal. In this state, the camera unit activates motion tracking mode, analyzing the deviation of the user's center of mass in the image sequence to determine the direction of departure. The infrared detection unit simultaneously monitors the changes in the user's foot contact point with the ground, generating three-dimensional motion trajectory data. For example, if the user exits the vehicle and moves toward the rear, the system compares the positional deviation of the user's contour in 10 consecutive image frames to calculate a movement speed of 0.8 meters per second and a deviation angle of 120 degrees. The detection module aligns the timestamps of these multi-source data (infrared thermal images, visible light video streams, door status, and seat pressure values) and then feeds them into a multimodal trajectory analysis model, generating trajectory prediction data that includes the user's spatial coordinates, movement vectors, and behavioral intentions. In particular, in complex environments (such as rainy nights with low visibility), the system prioritizes infrared image data for trajectory estimation. When the ambient light intensity drops below 10 lux, the system automatically increases the infrared sensor sampling frequency to 60 Hz and activates the camera unit's night vision fill light function to ensure the reliability of multi-source data collection.During the entire process, the near-field communication unit continuously monitors the signal strength index (RSSI). When the user carrying the portable vehicle control module moves to the communication boundary (about 8 meters), a communication link interruption warning is triggered. At this time, the detection module will switch to local data processing mode and execute the trajectory extrapolation algorithm based on the historical data of the last 10 seconds until the user completely leaves the monitoring area.

[0026] In some embodiments, the detection module includes an infrared detection unit, a near-field communication unit, and a vehicle docking unit. The near-field communication unit is configured to connect to the portable vehicle control module. The near-field communication unit is configured to cause the detection module to enter a user proximity detection state upon establishing a communication connection with the portable vehicle control module.

[0027] The near-field communication module is also configured to cause the detection module to terminate the user-away detection state upon disconnection from the portable vehicle control module. When in either the user-approach detection state or the user-away detection state, the infrared detection unit is configured to capture infrared image data of the user entering or exiting the vehicle. When either the user-approach detection state or the user-away detection state is in effect, the vehicle docking unit is configured to capture vehicle hardware control information.

[0028] Exemplarily, the detection module achieves precise perception of user behavior through a multi-unit collaborative mechanism. The near-field communication unit uses the Bluetooth Low Energy protocol to establish two-way communication with the portable vehicle control module. When a user carrying a smart key or a linked mobile phone enters the communication coverage range (typically 5 meters), the unit triggers a handshake protocol based on a signal strength threshold. For example, when a user approaches the digital key within 2 meters, the unit detects a continuously increasing RSSI signal value and immediately completes identity authentication with the key's embedded encryption chip, triggering the detection module to switch to user proximity detection. At this point, the infrared detection unit activates its highly sensitive thermal imaging sensor, capturing the thermal radiation distribution of the user's legs and feet at a rate of 25 frames per second. This generates a 160×120 resolution infrared image sequence, which is particularly suitable for identifying the user's approach direction at night or in low-light environments. The vehicle docking unit simultaneously collects door lock status, seat pressure sensor values, and electronic parking system status via the vehicle's CAN bus interface, forming a multi-dimensional hardware data stream. After the user boards the vehicle and leaves, the NFC unit determines the user's departure based on the signal strength decay rate. For example, if the signal strength drops below -70dBm for three consecutive seconds, the communication link is disconnected, triggering the detection module to terminate the user departure detection state. During this state, the vehicle docking unit continuously monitors the door angle sensor data. If the driver's door is open for more than 60 degrees for more than two seconds, the infrared detection unit is activated to dynamically track the foot area, analyzing the heat source movement trajectory in the continuous infrared image to determine the user's departure direction. In complex scenarios (such as multiple people approaching the vehicle at the same time), the system accurately identifies the target user by combining the NFC unit's connection priority determination with the spatial positioning data of the infrared image. For example, if two users with smart keys are located on the left and right sides of the vehicle, the system prioritizes the ambient lighting on the side with higher signal strength based on the difference in key signal strength and the heat source distribution characteristics in the infrared image. During the entire process, the vehicle docking unit synchronizes the transmission gear status and the door lock signal in real time. When it detects that the vehicle is in parking gear and the main driver's door is open, it automatically increases the sampling frequency of the infrared detection unit to 40Hz to capture fast-moving vehicle exit actions.

[0029] In some embodiments, the detection module further includes a camera unit mounted on the vehicle body. When in the user approach detection state or the user distance detection state, the camera unit is configured to capture visible light images of the user as they enter or exit the vehicle. The visible light images, hardware control information, and infrared image data constitute the entry and exit data.

[0030] In some embodiments, the detection module is installed with a trained multimodal trajectory analysis model, which includes: an image processing network and a text processing network, a fully connected network and an output network. When the detection module is used to generate trajectory prediction data based on boarding and disembarking data, it is specifically used to execute: S101-S105.

[0031] S101: Classify the boarding and alighting data to obtain image data and text data.

[0032] Exemplarily, the detection module receives raw vehicle entry and exit data from multiple sensors and then performs data classification. Image data includes thermal imaging sequences captured by the infrared detection unit, visible light video streams collected by the camera unit, and environmental images provided by the vehicle's surround-view camera. Text data includes hardware status information transmitted by the vehicle docking unit (such as door opening and closing angles, seat pressure values, and transmission gear codes), as well as device interaction logs recorded by the near-field communication unit. During the classification process, the system automatically identifies data based on packet header identifiers and content features: binary streams containing pixel matrix structures (such as H.264-encoded video frames with a resolution of 1920×1080) are classified as image data. Structured strings encoded in ASCII or Unicode are classified as text data. For example, when a user approaches a vehicle with a smart key, the system stores a 15-second video clip captured by the wide-angle camera at the bottom of the door and status messages transmitted via the CAN bus into the image data queue and text data buffer, respectively. The data classification module also performs timestamp alignment to ensure cross-modal data synchronization, such as the text record associated with the 120th frame of the video.

[0033] S102: Input the image data into an image processing network to obtain a three-dimensional model of the surrounding environment of the car, and obtain a feature vector of the user's movement trajectory based on the three-dimensional model.

[0034] Exemplarily, the image processing network uses a cascaded convolutional neural network architecture to process multi-source image inputs. For infrared thermal imaging data, the network first enhances the features of human heat source regions using a spatial attention mechanism. It then uses a three-dimensional convolution kernel to extract motion correlations between consecutive frames, generating a spatial displacement vector of the user's limbs. For visible light images, the network performs semantic segmentation to distinguish ground obstacles (such as accumulated water and gravel) from traversable areas, and reconstructs a depth map of the environment based on a stereo vision algorithm. The 3D modeling process integrates multi-view image features and generates a 3D scene model within a 5-meter radius of the vehicle using point cloud registration technology. For example, in rainy scenes, the system analyzes the segmentation results of reflective areas in the visible light image to mark slippery ground areas in the 3D model. Simultaneously, based on the movement trajectory of the user's foot heat source in the infrared image, the system calculates the user's position sequence in a polar coordinate system with the vehicle as the origin (e.g., the coordinates at time t1 (2.3m, 45°)), forming a 256-dimensional motion trajectory feature vector. This vector encodes key parameters such as the user's movement speed, directional stability, and path curvature.

[0035] S103: Input the text data into a text processing network to obtain a feature vector of the user's action intention.

[0036] For example, the text processing network extracts entities and models relationships from the raw text, mapping them into numerical feature vectors or converting them into state transition event markers. Using a multi-head attention mechanism, the network establishes cross-timestep correlations between text features, such as linking event chains, to infer the behavioral intent of "the user is getting on the bus." For complex scenarios (such as detecting multiple doors opening simultaneously), the network employs a graph neural network to reason about multi-entity relationships, generating a 128-dimensional action intention feature vector. This vector includes the user's goal-directed behavior (e.g., the strength of the intention to get on / off the bus), a coherence score for the action, and environmental constraint parameters (e.g., the obstacle warning level).

[0037] S104: Input the action trajectory feature vector and the action intention feature vector into a fully connected network to extract hidden features and obtain a fused hidden feature vector.

[0038] For example, a fully connected network concatenates a 256-dimensional motion trajectory feature vector with a 128-dimensional motion intention feature vector, then performs a nonlinear transformation through three hidden layers. The first hidden layer uses a channel-wise attention mechanism to dynamically weight motion trajectory features, strengthening the feature representation of key displacement nodes. The second hidden layer employs gated recurrent units to capture temporal dependencies and establish an association model between the user's historical trajectory and current intention. The third hidden layer performs feature compression, reducing the 1024-dimensional intermediate features to a 512-dimensional fused hidden feature vector. For example, when a user approaches a vehicle along a curved path, the network increases the weight of the lateral displacement feature and, combined with the intent features in the text data, emphasizes the predicted trajectory tendency in the driving direction in the fused vector. A contrastive learning strategy is employed during network training, using positive and negative sample pairs (e.g., straight-line approach vs. roundabout approach) to enhance feature discrimination.

[0039] S105: Input the fused hidden feature vector into the output network, and obtain trajectory prediction data through activation function and loss function.

[0040] For example, the input-output network feeds the fused hidden feature vector into a parallel branch structure. The spatial coordinate branch uses a spatial transformation layer to output a Cartesian coordinate sequence (e.g., [(x1, y1), (x2, y2)]) of the user's location within the next three seconds. The behavior prediction branch uses a softmax activation function to calculate the probability distribution of behaviors such as stopping, turning, and accelerating. The environmental interaction branch generates vehicle response recommendations (e.g., pre-activating the corresponding side ambient light). The loss function combines mean squared error (coordinate prediction), cross entropy (behavior classification), and policy gradient (vehicle response) for multi-objective optimization. For example, if the fused features indicate that the user is approaching the passenger door at a speed of 1.2 m / s at a 30-degree angle, the network output includes a coordinate prediction for five time steps, a high-probability "straight line" behavior judgment, and a vehicle response command to trigger the right welcome light carpet. The output data is then smoothed using a Kalman filter to eliminate trajectory jitter caused by sensor noise.

[0041] In some embodiments, the preset control instruction library includes: welcome start instructions, carpet guidance instructions, safety warning instructions, lateral following instructions, energy-saving maintenance instructions and light carpet fade-out instructions. When the instruction generation module is used to match the preset control instruction library according to the trajectory prediction data and obtain the atmosphere light control instruction, it is specifically used to execute: S201-S208.

[0042] S201: Obtain the user's real-time coordinates, moving speed vector, and horizontal projection distance from the vehicle chassis in trajectory prediction data.

[0043] For example, a dynamic perception baseline is established by acquiring the user's spatial coordinates, velocity vector, and horizontal projection distance in real time. Multi-dimensional data fusion ensures centimeter-level positioning accuracy, and the ability to resolve velocity vector direction supports the prediction of user movement trends. Horizontal projection distance calculation uses vehicle coordinate system transformation to eliminate errors caused by terrain fluctuations, providing reliable spatial parameters for subsequent command matching.

[0044] S202 : Divide the range into a first distance threshold interval, a second distance threshold interval, and a third distance threshold interval based on the horizontal projection distance.

[0045] For example, a scenario-based hierarchical response mechanism is implemented by dividing the distance threshold intervals into three ranges: the first range is 3-5 meters, the second range is 1-3 meters, and the third range is 0-1 meter. The 3-5 meter range corresponds to the mid-range welcoming scenario, the 1-3 meter range handles the need for precise path guidance, and the 0-1 meter range enhances safety protection. The dynamic interval boundary setting takes into account ergonomics and optical projection characteristics to avoid user experience fragmentation caused by premature or delayed responses.

[0046] S203: When it is detected that the user is in the first distance threshold interval and the moving speed vector is pointing to the vehicle, match the welcome start instruction.

[0047] For example, a welcome start command controls the ambient lighting to project a first-type virtual carpet in a breathing gradient pattern. This first-type virtual carpet includes dynamic welcome text and a radial light diffusion effect. The welcome start command triggers active vehicle interaction, and the breathing gradient light effect reduces the irritation of bright nighttime light. The dynamic welcome text projection enhances brand recognition, and the radial light diffusion creates a visual guidance channel. This mode enhances the sense of ceremony through contactless interaction, and the light carpet coverage is matched to the distance threshold to avoid inefficient energy consumption.

[0048] S204 : When the user enters the second distance threshold interval, matching the carpet guidance instruction based on the direction angle deviation value of the moving speed vector.

[0049] For example, if the detected angular deviation is less than 15 degrees, the linear guidance mode is activated, projecting a strip of light blanket instructions that coincide with the door trajectory. If the detected angular deviation is greater than or equal to 15 degrees, the path correction mode is activated, projecting a fan-shaped light blanket instruction with directional arrows. In the dual-mode guidance mechanism, the linear guidance mode uses a strip of light blanket to enhance visual guidance, with a 15-degree tolerance range to adapt to natural walking fluctuations. The dynamic arrow pointing feature of the fan-shaped light blanket can correct for path deviations of up to 45 degrees, and the beam angle adjusts adaptively with the deviation value.

[0050] S205: When it is detected that the user enters a third distance threshold interval, a safety warning instruction is matched.

[0051] For example, the safety warning system multiplies the brightness of the vehicle's luminance to enhance visibility in hazardous areas. A ring-shaped aperture forms a dynamic safety boundary, and pulse-modulated brightness prevents glare. The projection area shrinks and expands in real time with the user's gait. Specifically, the ambient lighting is controlled to increase the brightness to 1500 lumens, and a ring-shaped warning aperture is superimposed on the projection area at the user's feet. This mechanism effectively prevents collisions when getting on and off the vehicle, providing proactive warnings for obstacles in blind spots.

[0052] S206: If it is detected that the user is in a lateral movement trajectory, match the lateral follow instruction.

[0053] For example, lateral tracking commands enable dynamic spatial registration of the light carpet projection, causing the projection angle of the virtual carpet to deflect in real time as the user moves, and generating a flowing light effect at the carpet's edge to follow the user's movements. This flowing light effect uses a phase synchronization algorithm, with displacement delay controlled within 200ms. This mode maintains continuous visual guidance, resolves the problem of light carpet interruption when the user moves around the vehicle, and enhances the system's smooth tracking.

[0054] S207: When it is detected that the user's inactivity time exceeds a preset threshold, match the energy-saving maintenance instruction.

[0055] For example, the energy-saving maintenance command reduces power consumption by up to 70% through the coupled control of brightness attenuation and light field contraction. It uses thermal imaging to locate the core area where users stand, maintaining the minimum necessary lighting. Specifically, the virtual carpet brightness is reduced to 30% of the standard value, and the light carpet is controlled to contract to the user's standing area. This mode extends device life while ensuring basic functionality, and is particularly suitable for high-frequency usage scenarios such as ride-hailing.

[0056] S208 , matching the light carpet fading instruction in stages according to the detected trajectory of the user when getting off the vehicle.

[0057] For example, the fade sequence uses an exponential decay curve to avoid visual abruptness, the retraction animation dynamically matches the user's moving away speed, and the distance-light intensity mapping relationship conforms to the Weber-Fechner law. When the preset distance redundancy is reached, frequent starts and stops caused by false triggering are prevented. Specifically, when the horizontal projection distance reaches 2 meters, the fade sequence command is activated, the distance reaches 3 meters, the light carpet retraction animation is triggered, and the virtual carpet projection command is turned off when the distance exceeds 5 meters.

[0058] In some embodiments, the atmosphere light control module is used to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction, specifically to execute: S301-S302.

[0059] S301: parsing projection mode parameters in the ambient light control instruction, and dynamically adjusting the projection angles of multiple ambient lights.

[0060] For example, dynamic projection angle adjustment is based on a vector relationship between the vehicle's four-wheel alignment sensor data and the user's real-time coordinates. When an angle is detected between the user's movement direction and the centerline of the vehicle door (e.g., a user approaching the passenger door at a 45-degree angle), the module acquires the door hinge angle data via the CAN bus and calculates the required compensation deflection for the ambient light on both sides. When an angle is detected between the user's movement direction and the door opening direction, the ambient light on both sides is controlled to rotate differentially to compensate for the projection angle offset. For example, if a user approaches from the right rear of the vehicle at a 30-degree deviation angle, the right-hand light cluster maintains its baseline projection direction, while the left-hand light cluster rotates clockwise at a differential speed of 15 revolutions per minute, shifting the intersection of the two light beams 1.2 meters forward along the user's path. Differential control utilizes PID closed-loop regulation, updating the pulse frequency of the rotary stepper motor every 200ms to ensure that the light carpet centerline never deviates from the user's trajectory by more than 5 centimeters. In complex terrain scenarios (such as slopes), the module combines pitch angle data from the Inertial Measurement Unit (IMU) to dynamically adjust the projection angle. For example, if the vehicle leans forward 3 degrees, the module automatically increases the headlight tilt angle to maintain a horizontal projection of the light blanket. The projection parameter memory function stores commonly used configurations. When repetitive user movement patterns are detected (such as a daily approach to a parking space), the module pre-sets the headlight deflection angle, reducing response latency to less than 80ms.

[0061] S302: According to the light intensity coefficient carried in the ambient light control instruction, adjust the driving current of the light-emitting unit of the ambient light, and control the light-emitting unit to emit light according to a preset brightness gradient.

[0062] For example, sinusoidal dimming in welcome mode periodically varies the drive current at a 2Hz frequency, allowing the light intensity to fluctuate smoothly between 300 and 1200 lumens, simulating a natural breathing rhythm. For example, in a nighttime welcome scenario, if the module detects ambient light intensity below 50 lux, it automatically increases the baseline brightness by 20% while maintaining the sinusoidal dimming curve to ensure visibility and prevent glare. When safety alert mode is activated, the driver circuit switches to direct digital synthesis (DDS) mode, increasing the current to 120% of the rated value within 5ms, achieving a transient high-intensity light output of 1500 lumens. This mode incorporates dynamic thermal management. When the temperature sensor detects that the LED substrate exceeds 75°C, it automatically inserts a 10ms intermittent power-off period to maintain a safe operating temperature. For special weather conditions (such as rain and fog), the module triggers a penetration enhancement mode based on a rain sensor signal, adjusting the color temperature from 6500K to 3000K while increasing the red light spectrum intensity to 40%. Brightness gradient control also includes a scene-adaptive compensation algorithm. When the user is wearing dark clothing, the lumen output is automatically increased by 20% based on the clothing reflectivity data fed back by the camera to ensure that the light carpet pattern is clearly visible.

[0063] S303: Synchronously control the coverage of the virtual carpet based on the vehicle door lock status and the door opening and closing angle.

[0064] For example, when a door opening angle greater than 60 degrees is detected, wide-area coverage mode is activated, expanding the light carpet width to 1.2 times the standard width and generating a dynamically projected boundary marker at the bottom of the door. Specifically, when a door opening angle exceeds the 60-degree threshold (e.g., when a user pulls the door wide open to load something), the module activates wide-area coverage mode: by adjusting the deflection angle of the digital micromirror device (DMD), the light carpet is laterally expanded to 1.2 times the standard width, while a dynamic dashed boundary marker is projected along the bottom edge of the door. For example, in a seven-seat vehicle, when the sliding door opens at a 70-degree angle, the light carpet width expands from the default 0.8 meter to 0.96 meters, covering a larger footrest area, and an alternating yellow warning stripe is generated at the door track. This mode incorporates terrain perception data, automatically overlaying a red diagonal marker at the edge of the expanded light carpet when the ultrasonic sensor detects a ground drop exceeding 5 centimeters. The dynamically projected boundary uses frame-by-frame rendering technology, adjusting the refresh rate of the sign in real time based on the door's movement speed (for example, if the door opens at 10 degrees per second, the boundary sign refreshes at 25Hz), ensuring visual continuity. In rainy weather, the module uses data from the rain sensor to switch the boundary sign to a high-contrast blue wavy pattern, enhancing the slippery road warning effect.

[0065] S304: Execute dynamic synchronous control of light effects, using PWM pulse modulation to make the light beams of multiple atmosphere lights produce interference effects, forming periodic flowing light patterns on the surface of the virtual carpet, wherein the propagation direction of the flowing light patterns is synchronized in the opposite direction to the user's movement speed vector.

[0066] For example, the dynamic lighting effect synchronization control module utilizes a master-slave clock synchronization architecture, sending PWM modulation signals with a phase difference of π / 3 to each ambient light fixture via the I2C bus. Each fixture's driver chip incorporates a built-in 32-bit timer, precisely controlling the on / off timing of the LED array, resulting in multiple light beams producing alternating light and dark interference fringes on the projection plane. For example, when three ambient light clusters operate with a 120-degree phase difference, a flowing light pattern with a wavelength of approximately 15 cm forms on the virtual carpet surface. The propagation speed is inversely correlated with the user's movement speed: when the user moves forward at 1 m / s, the light pattern flows in the opposite direction at 0.8 m / s, creating a dynamic visual guidance effect. The interference pattern generator receives real-time vehicle posture data collected by the IMU. When detecting vehicle tilt (e.g., parked on a 5-degree slope), it automatically compensates for geometric distortion in the light pattern propagation direction. In scenarios where multiple people are boarding and exiting simultaneously, the system independently assigns light pattern parameters to each user, generating, for example, longitudinal ripples for the driver and radial ripples for rear passengers. Frequency division multiplexing (FDM) is used to prevent overlapping light effects. The color temperature of the flowing light pattern is automatically adjusted according to the ambient light. 5500K cold white light is used during the day to improve contrast, and it switches to 4000K warm white light at night to reduce visual fatigue.

[0067] See also Figure 2 , Figure 2 1 is a schematic block diagram of a car with carpet atmosphere lighting provided by an embodiment of the present application. Car 200 with carpet atmosphere lighting is equipped with a carpet atmosphere lighting as described in the embodiment of the present application. Car 200 with carpet atmosphere lighting can be connected to a server to assist in implementing the computing functions to be implemented in the embodiment of the present application.

[0068] Among them, the server can be an independent server, a server cluster, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0069] like Figure 2 As shown, a car 200 with carpet ambient light includes: an ambient light 201 , a detection module 202 , a command generation module 203 and an ambient light control module 204 .

[0070] Ambient lights 201 are installed on both sides of the car chassis to generate a virtual carpet; Detection module 202, used to obtain the user's boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data; The instruction generation module 203 is used to receive the trajectory prediction data and match it with a preset control instruction library according to the trajectory prediction data to obtain the atmosphere light control instruction; The atmosphere light control module 204 is configured to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction.

[0071] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A carpet atmosphere lamp, characterized in that: The carpet atmosphere lamp includes: Ambient lights, which are installed on both sides of the car chassis and are used to generate a virtual carpet; A detection module, configured to obtain the user's boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data; An instruction generation module is used to receive the trajectory prediction data and match it with a preset control instruction library according to the trajectory prediction data to obtain an atmosphere light control instruction; The atmosphere light control module is used to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction.

2. The carpet atmosphere lamp according to claim 1, characterized in that: The detection module is further configured to be in communication with the user's portable vehicle control module; When the detection module receives the start control signal from the portable vehicle control module, it enters the user approach detection state and obtains the user boarding data; When the detection module detects that the user leaves the car, it enters the user leaving detection state and obtains the user's getting off data.

3. The carpet atmosphere lamp according to claim 2, characterized in that: The detection module includes: an infrared detection unit, a near field communication unit and a vehicle docking unit, wherein the near field communication unit is used to connect with the portable vehicle control module; The near field communication unit is used to make the detection module enter the user approach detection state when completing the communication connection with the portable vehicle control module; The near field communication module is further configured to enable the detection module to terminate the user away detection state when the communication with the portable vehicle control module is disconnected; When in the user approach detection state or the user distance detection state, the infrared detection unit is used to obtain infrared image data of the user when getting on or off the vehicle; When in the user approach detection state or the user distance detection state, the vehicle docking unit is used to obtain hardware control information of the vehicle.

4. The carpet atmosphere lamp according to claim 3, characterized in that: The detection module further includes: a camera unit, which is used to be installed on the vehicle body; wherein, When in the user approach detection state or the user distance detection state, the camera unit is used to obtain visible light images of the user when getting on and off the vehicle, and the visible light images, the hardware control information and the infrared image data constitute the getting on and off the vehicle data.

5. The carpet atmosphere lamp according to claim 1, wherein: The detection module is equipped with a trained multimodal trajectory analysis model, which includes an image processing network, a text processing network, a fully connected network, and an output network. When the detection module is used to generate trajectory prediction data based on the boarding and alighting data, it is specifically used to perform the following: Classifying the boarding and alighting data to obtain image data and text data; Inputting the image data into the image processing network to obtain a three-dimensional model of the surrounding environment of the car, and obtaining a user's movement trajectory feature vector based on the three-dimensional model; Inputting the text data into the text processing network to obtain a user's action intention feature vector; Inputting the action trajectory feature vector and the action intention feature vector into the fully connected network to extract hidden features and obtain a fused hidden feature vector; The fused hidden feature vector is input into the output network, and the trajectory prediction data is obtained through the activation function and the loss function.

6. The carpet atmosphere lamp according to claim 3, characterized in that: The preset control instruction library includes: a welcome start instruction, a carpet guidance instruction, a safety warning instruction, a lateral follow instruction, an energy-saving maintenance instruction, and a light carpet fading instruction. When the instruction generation module is used to execute the matching of the preset control instruction library according to the trajectory prediction data to obtain the ambient light control instruction, it is specifically used to execute: Obtaining the user's real-time coordinates, moving speed vector, and horizontal projection distance from the vehicle chassis in the trajectory prediction data; Dividing the distance into a first distance threshold interval, a second distance threshold interval, and a third distance threshold interval based on the horizontal projection distance; When it is detected that the user is in the first distance threshold interval and the moving speed vector is pointing to the vehicle, matching the welcome start instruction; When the user enters the second distance threshold interval, matching the carpet guidance instruction based on the direction angle deviation value of the moving speed vector; When detecting that the user enters the third distance threshold interval, matching the safety warning instruction; If it is detected that the user is in a lateral movement trajectory, matching the lateral follow instruction; When it is detected that the user's inactivity time exceeds a preset threshold, matching the energy-saving maintenance instruction; According to the detected trajectory of the user when getting off the vehicle, the light carpet fading instruction is matched in stages.

7. The carpet atmosphere lamp according to claim 1, characterized in that: The atmosphere light control module is used to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction, specifically to perform: parsing the projection mode parameters in the atmosphere light control command, and dynamically adjusting the projection angles of the plurality of atmosphere lights; According to the light intensity coefficient carried in the atmosphere light control instruction, adjusting the driving current of the light-emitting unit of the atmosphere light, and controlling the light-emitting unit to emit light according to a preset brightness gradient; Synchronously controlling the coverage of the virtual carpet based on the vehicle door lock status and the door opening and closing angle; The light beams of the plurality of atmosphere lights are caused to produce an interference effect through PWM pulse modulation, thereby forming a periodic flowing light pattern on the surface of the virtual carpet.

8. A car with carpet atmosphere lighting, characterized in that: A car with a carpet atmosphere light is installed with the carpet atmosphere light according to any one of claims 1 to 7, and the car with a carpet atmosphere light comprises: Ambient lights, which are installed on both sides of the car chassis and are used to generate a virtual carpet; A detection module, configured to obtain the user's boarding and alighting data and generate trajectory prediction data based on the boarding and alighting data; An instruction generation module is used to receive the trajectory prediction data and match it with a preset control instruction library according to the trajectory prediction data to obtain an atmosphere light control instruction; The atmosphere light control module is used to control the atmosphere light to generate a virtual carpet according to the atmosphere light control instruction.

Images