Bicycle backward sensing method and system based on multiple radars and intelligent glasses

Abstract

The invention discloses a bicycle backward sensing method and system based on multiple radars and intelligent glasses, and the method comprises the steps: obtaining the echo data of a rear millimeter wave radar and a front millimeter wave radar, and obtaining a radar measurement sequence; performing target detection and parameter estimation on the radar measurement sequence, outputting a target measurement set containing a target distance and an azimuth angle, and organizing the target measurement set under a vehicle body coordinate system; performing cross-radar data association and shielding recovery on the target measurement set from the front millimeter-wave radar and the rear millimeter-wave radar, and generating a target track sequence covering a 180-degree range behind the vehicle body; acquiring a head posture of the intelligent glasses, converting the target track sequence from a vehicle body coordinate system to a head coordinate system according to pre-calibrated vehicle body and head external parameters, and generating overlapping display data consistent with a target orientation and marked with a distance; and rendering the superimposed display data in the field of view of the intelligent glasses. According to the technical scheme, the riding safety is improved.

Description

Technical Field

[0001] This invention relates to the field of smart glasses technology, and in particular to a bicycle rearward perception method and system based on multiple radars and smart glasses. Background Technology

[0002] With the advancement of millimeter-wave radar technology towards lighter weight and lower power consumption, some high-end bicycles have begun to integrate rear-facing radar to detect vehicles approaching from behind, thereby enhancing riding safety. However, existing technologies mostly use a single rear-facing radar to achieve coarse-grained distance detection and alert riders to approaching vehicles via a cycling computer or mobile terminal. They lack accurate reconstruction of the dynamic changes in the position and angle of rear targets, which can easily lead to target trajectory loss and false alarms, especially in high-speed close-up or multi-vehicle parallel scenarios. Summary of the Invention

[0003] The purpose of this application is to propose a bicycle rearward perception method and system based on multiple radars and smart glasses, so as to solve the technical problems of insufficient timeliness, stability and reliability of radar prompts.

[0004] To address the aforementioned technical problems, this application provides a bicycle rearward perception method based on multiple radars and smart glasses, employing the following technical solution: A bicycle rearward perception method based on multiple radars and smart glasses includes the following steps: The echo data of the rear millimeter-wave radar located at the rear of the vehicle body and the front millimeter-wave radar located at the front of the vehicle body are acquired and time-synchronized with the vehicle attitude data to obtain the radar measurement sequence. Target detection and parameter estimation are performed on the radar measurement sequence, and a target measurement set containing target distance and azimuth is output. The target measurement set is then organized in the vehicle coordinate system. Perform cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body; The head posture of the smart glasses is obtained, and the target trajectory sequence is transformed from the vehicle coordinate system to the head coordinate system based on the pre-calibrated vehicle and head extrinsic parameters to generate superimposed display data that is consistent with the target orientation and marked with distance. The overlay display data is rendered within the field of view of the smart glasses.

[0005] In one possible implementation, before the step of performing cross-radar data correlation and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body, the method further includes: The target measurement data from the front and rear millimeter-wave radars are unified into the vehicle coordinate system and time-aligned according to the synchronization timestamp. Calculate the distance difference, azimuth difference, and relative velocity difference separately, and set multidimensional gating thresholds to eliminate measurements that exceed the thresholds; Preserve the same frame candidate target measurement set used for cross-radar data correlation.

[0006] In one possible implementation, in the step of performing cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle, the radar data association step specifically includes: Construct a joint cost function weighted by relative velocity continuity, azimuth neighborhood overlap, and echo amplitude consistency; The minimum cost matching algorithm is used to generate cross-radar matching pairs in the same frame candidate target measurement set; Perform state estimation and update on the matching pairs, and output target trajectory parameters with trajectory labels; Unmatched measurements are marked as new trajectories or temporary trajectories.

[0007] In one possible implementation, in the step of performing cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body, the occlusion recovery step specifically includes: If the rear millimeter-wave radar fails to generate an effective measurement of the existing trajectory in consecutive preset frames or its echo amplitude is lower than the obstruction criterion threshold, motion prediction is performed based on the trajectory parameters of the previous moment to obtain the predicted position. Search for targets matching the predicted position in the candidate measurements of the front millimeter-wave radar and perform trajectory completion. When the completion is successful, update the trajectory parameters and keep the trajectory identifier continuous. When the completion fails, set the trajectory to the observation state until effective observation is obtained or the termination threshold is reached. The updated trajectory is subjected to duplicate target merging and angle conflict resolution, and the output is limited to the target trajectory sequence within a 180° range behind the vehicle body.

[0008] In one possible implementation, the step of acquiring the head posture of the smart glasses, transforming the target trajectory sequence from the vehicle coordinate system to the head coordinate system based on pre-calibrated vehicle and head extrinsic parameters, and generating overlay display data consistent with the target orientation and labeled with distance, specifically includes: Based on the preset spatial extrinsic parameters between the vehicle body and the head and the currently acquired head posture information, coordinate transformation is performed on the position and azimuth of each target trajectory to obtain the target position and direction with the head coordinate system as a reference. Based on the coordinate transformation results, construct the corresponding graphic annotation content, output the superimposed display data containing the target's true orientation and distance, and keep the target graphic's position in the field of view consistent with the target's orientation in physical space.

[0009] In one possible implementation, after the step of performing coordinate transformation on the position and azimuth of each target trajectory based on preset spatial extrinsic parameters between the vehicle body and the head and the currently acquired head posture information to obtain the target position and orientation with the head coordinate system as a reference, the method further includes: The azimuth angle estimate of the target trajectory is filtered and the azimuth angle fluctuation range of consecutive frames is calculated; When the fluctuation range does not exceed the stability threshold and the synchronization error between the vehicle body and the head is less than the time synchronization threshold, the target is fixed in the corresponding direction within the field of view of the smart glasses by using the line-of-sight locking method.

[0010] In one possible implementation, the step of acquiring echo data from the rear millimeter-wave radar located at the rear of the vehicle and the front millimeter-wave radar located at the front of the vehicle, and combining this with vehicle attitude data for time synchronization to obtain a radar measurement sequence, specifically includes: The vehicle controls the front and rear millimeter-wave radars to periodically transmit and receive echo signals in a time-division alternating manner. Preliminary pulse compression and frequency offset estimation are performed on the echo signals respectively; The timestamp difference is calculated based on the vehicle attitude data, and the measurement alignment is completed to output a radar measurement sequence with a synchronization timestamp.

[0011] To address the aforementioned technical problems, this application also provides a bicycle rearward perception system based on multiple radars and smart glasses, employing the following technical solution: A bicycle rearward perception system based on multiple radars and smart glasses includes: The acquisition module is used to acquire the echo data of the rear millimeter-wave radar located at the rear of the vehicle body and the front millimeter-wave radar located at the front of the vehicle body, and to combine the vehicle attitude data for time synchronization to obtain the radar measurement sequence. The output module is used to perform target detection and parameter estimation on the radar measurement sequence, output a target measurement set including target distance and azimuth, and organize the target measurement set in the vehicle coordinate system; The execution module is used to perform cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar, and generate a target trajectory sequence covering a 180° range behind the vehicle body; The conversion module is used to obtain the head posture of the smart glasses, and convert the target trajectory sequence from the vehicle coordinate system to the head coordinate system according to the pre-calibrated vehicle and head extrinsic parameters, and generate superimposed display data that is consistent with the target orientation and marked with distance; The display module is used to render the overlay display data within the field of view of the smart glasses.

[0012] To address the aforementioned technical problems, this application also provides a computer device that employs the following technical solution: A computer device includes a memory and a processor, the memory storing computer-readable instructions, the processor executing the computer-readable instructions to implement the steps of the bicycle rearward perception method based on multiple radars and smart glasses as described above.

[0013] To address the aforementioned technical problems, this application also provides a computer-readable storage medium, employing the technical solution described below: A computer-readable storage medium storing computer-readable instructions, which, when executed by a processor, implement the steps of the bicycle rearward perception method based on multiple radars and smart glasses as described above.

[0014] Compared with the prior art, the embodiments of this application have the following main advantages: The bicycle rearward perception method disclosed in this application, based on multiple radars and smart glasses, completes detection, association, and occlusion recovery using front and rear millimeter-wave radars under a unified time base. It also transforms the trajectory from the vehicle coordinate system to the head coordinate system for consistent line-of-sight display, providing continuous, stable, and true-direction target presentation within a 180° range behind the vehicle. This significantly reduces false alarms and missed alarms, shortens rider decision-making and avoidance time, and improves riding safety. Attached Figure Description

[0015] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of an embodiment of a bicycle rearward perception method based on multiple radars and smart glasses according to this application; Figure 2 This is a schematic diagram of a structure of an embodiment of a bicycle rearward perception system based on multiple radars and smart glasses according to this application; Figure 3This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0018] refer to Figure 1 The diagram illustrates a flowchart of an embodiment of a bicycle rearward perception method based on multiple radars and smart glasses according to this application. The bicycle rearward perception method based on multiple radars and smart glasses includes the following steps: It should be noted that millimeter-wave radar refers to radar sensors operating in the 24 GHz or 76–81 GHz frequency band; azimuth refers to the horizontal rotation angle of the target relative to the vehicle coordinate system; the vehicle coordinate system is defined as follows: X-axis points forward along the front of the vehicle, Y-axis points to the left side of the vehicle, and Z-axis points vertically upward; the head coordinate system is defined with the optical axis of the smart glasses as the forward direction, and the horizontal rightward and vertical upward directions of the glasses are the horizontal and vertical directions, respectively. Extrinsic parameters are the rigid body transformation parameters (R, t) from the vehicle coordinate system to the head coordinate system, used for spatial mapping between different carriers.

[0019] Step S101: Obtain echo data from the rear millimeter-wave radar located at the rear of the vehicle body and the front millimeter-wave radar located at the front of the vehicle body, and synchronize them with the vehicle attitude data to obtain the radar measurement sequence.

[0020] In this embodiment, the electronic device running on the bicycle rearward perception method based on multi-radar and smart glasses can send or receive data via wired or wireless connection. It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra-wideband) connections, and other currently known or future-developed wireless connection methods.

[0021] In this embodiment, echo data from millimeter-wave radars located at the rear and front of the vehicle body are first acquired and synchronized with vehicle attitude data (from the vehicle-mounted IMU and wheel speedometer) to form a radar measurement sequence. Time synchronization is preferably achieved through a unified time base and timestamp: the two radars alternately transmit / receive in a time-division manner, sending pilot signals for frequency offset estimation and writing a unified timestamp in each working cycle; the vehicle-mounted processor uses a high-precision timer shared with the IMU to stamp the arriving echo frames, and then fine-tunes them according to a linear drift model to ensure that the time difference between the two radars and the IMU does not exceed a predetermined threshold (e.g., ≤5 ms). For example, the radar PRF is set to 2–4 kHz, the measurement frame rate to 20–50 Hz, and the attitude sampling rate to ≥100 Hz; after synchronization, a time-ordered radar measurement sequence is obtained, serving as a unified input for subsequent detection and fusion.

[0022] Step S102: Target detection and parameter estimation are performed on the radar measurement sequence, and a target measurement set containing target distance and azimuth is output. The target measurement set is then organized in the vehicle coordinate system.

[0023] In this embodiment, target detection and parameter estimation are performed on the radar measurement sequence, outputting a target measurement set containing target range and azimuth, and organizing the data in the vehicle coordinate system. Detection can employ Constant False Alarm Rate (CFAR) to select candidate echoes on the range-Doppler plane. Parameter estimation uses range FFT and Doppler FFT to obtain range and relative velocity, which are then combined with digital beamforming and angle estimation algorithms (such as beam scanning of a MIMO virtual array or MUSIC / ESPRIT) to obtain the azimuth. For ease of subsequent fusion, each target is represented as {r, θ, v_rel, SNR}, where r is the radial range, θ is the horizontal azimuth (0° for the positive X-axis and positive for counterclockwise), v_rel is the relative velocity along the radar line of sight, and SNR is the signal-to-noise ratio. Polar coordinate measurements are converted to planar coordinates in the vehicle coordinate system for consistent processing. For example, the range resolution of a 24 GHz radar can be 0.6–0.9 m, and the angular resolution depends on the antenna array (typically 2–5°). By using a virtual array synthesized from the left and right subarrays, the azimuth resolution can be improved to the order of 2° to adapt to close-range vehicle passing scenarios.

[0024] Step S103: Perform cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body.

[0025] In this embodiment, after obtaining the target measurement sets for the front and rear, cross-radar data association and occlusion recovery are performed to generate a target trajectory sequence covering a 180° range behind the vehicle. To ensure matching of the two measurements within the same frame, they are first aligned according to the synchronization timestamp within the same time window and the physical quantities are normalized. Then, gating conditions (such as |Δr|≤r_gate, |Δθ|≤θ_gate, |Δv|≤v_gate) are set to eliminate out-of-bounds candidates. A joint cost function J=w_r|Δr|+w_θ|Δθ|+w_v|Δv|+w_a|Δamp| is constructed on the gating results, and Hungarian matching is used to generate matching pairs. The trajectory status is updated using recursive filtering (such as α-β filtering or Kalman filtering), and unmatched measurements are managed as new or temporary targets. When the rear radar has no effective measurements for a certain established trajectory for several consecutive frames or the SNR is lower than the occlusion criterion, motion prediction is performed based on the previous state to obtain the predicted position, and a matching point is searched in the front radar measurement to fill the gap. If the match is successful, the trajectory label is maintained and the state is updated; otherwise, it enters the buffer until a new observation is obtained or the termination threshold is reached. To avoid duplicate labeling, overlap resolution and duplicate target merging are performed on the updated trajectory, and only the trajectory in the rear half-plane of the vehicle body (i.e., θ∈[90°, 270°]) is retained as the output. In actual implementation, the gating threshold can be adaptively adjusted according to vehicle speed and target density, for example, r_gate=2–6m, θ_gate=3–8°, v_gate=2–4 m / s, to balance recall and accuracy in congested and sparse scenarios.

[0026] Step S104: Obtain the head posture of the smart glasses, and transform the target trajectory sequence from the vehicle coordinate system to the head coordinate system according to the pre-calibrated vehicle and head extrinsic parameters to generate superimposed display data that is consistent with the target orientation and marked with distance.

[0027] In this embodiment, to achieve visual cues consistent with the real direction in smart glasses, the head posture of the smart glasses is acquired, and the target trajectory sequence is transformed from the vehicle coordinate system to the head coordinate system, generating superimposed display data consistent with the target orientation and labeled with distance. The head posture is provided by the IMU at the glasses (update rate ≥100 Hz), and the extrinsic parameters R and t are obtained through calibration after factory manufacturing or assembly (samples can be collected and solved in both static and uniform motion states). The vehicle-side or glasses-side processor calculates p_h=R·p_b+t based on this to map the target position in the vehicle system to the position in the head system. To ensure stable display of gaze lock, the converted azimuth sequence is smoothed in the temporal domain, and the azimuth fluctuation range and time synchronization error are evaluated. When the azimuth variance is lower than the stability threshold (e.g., ≤(2°)^2) and the duration exceeds the set duration (e.g., ≥200 ms), while the time synchronization error is less than the threshold, superimposed display data of gaze lock is generated; otherwise, it switches to a degenerate style (e.g., directional arrows at the edge of the field of view + simplified distance labels) and prompts for quick extrinsic parameter verification. Overlay data is preferably displayed using a circular sector to represent the orientation, a concentric circle to mark the distance, and a relative speed or approach time next to the target to improve readability.

[0028] Step S105: Render the overlay display data within the field of view of the smart glasses.

[0029] In this embodiment, rendering can be completed locally on the glasses or generated by the vehicle's computing unit and transmitted to the glasses via a low-latency wireless link. To meet the real-time requirements for riding safety, the measured end-to-end latency should be controlled within a predetermined upper limit (e.g., ≤80 ms). When the link quality deteriorates or the computational load increases, the rendering detail or update rate can be reduced to ensure that the latency does not exceed the limit. An implementable example is as follows: a rear 24 GHz radar is installed on the seatpost, and a front 24 GHz radar is installed on the stem, both working alternately at a 20 Hz frame rate; the vehicle's main controller updates the trajectory status at 50 Hz and sends targets within 180° behind to the smart glasses in polar coordinates of the head coordinate system; the glasses refresh the UI at 60 Hz, displaying a sector + distance circle line-of-sight lock when stability conditions are met, otherwise indicating with an edge arrow. The average latency of the entire link is measured to be 65 ms.

[0030] This application uses front and rear millimeter-wave radars to complete detection, correlation, and occlusion recovery under a unified time base, and transforms the trajectory from the vehicle coordinate system to the head coordinate system for consistent line-of-sight display. It can provide a continuous, stable target presentation within a 180° range behind the vehicle that is consistent with the actual direction, thereby significantly reducing false alarms and missed alarms, shortening the rider's decision-making and avoidance time, and improving riding safety.

[0031] In some optional implementations of this embodiment, before the step of performing cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body, the method further includes: The target measurement data from the front and rear millimeter-wave radars are unified into the vehicle coordinate system and time-aligned according to the synchronization timestamp. Calculate the distance difference, azimuth difference, and relative velocity difference separately, and set multidimensional gating thresholds to eliminate measurements that exceed the thresholds; Preserve the same frame candidate target measurement set used for cross-radar data correlation.

[0032] In this embodiment, the target measurement data output by the front and rear millimeter-wave radars is first unified to the vehicle coordinate system by transforming the vehicle coordinate system obtained from assembly or calibration to the coordinate systems of each millimeter-wave radar sensor. Time alignment uses a unified time base timestamp, and measurements falling within the same time window are grouped into candidates for the same frame. Subsequently, the distance difference, azimuth difference, and relative speed difference between targets within the same frame are calculated, and multi-dimensional gating thresholds are set to eliminate out-of-bounds measurements. The gating thresholds can be adaptively set according to vehicle speed or target density; for example, the distance difference gating is set to 2 to 6 meters, the azimuth difference gating is set to 3 to 8 degrees, and the relative speed difference gating is set to 2 to 4 meters per second. The candidate target measurement set retained after gating is used for subsequent matching and trajectory updates, which can significantly reduce mismatches and false alarms, and ensure that the correlation problem is solvable and stable.

[0033] This application aligns data in the vehicle coordinate system and under the synchronization timestamp, and implements multi-dimensional gating with distance difference, azimuth difference and relative speed difference. This effectively eliminates abnormal and noisy measurements, reduces erroneous matching and computational scale, and provides a stable, reliable and comparable set of candidate frames for cross-radar trajectory association, thereby improving the fusion quality from the source.

[0034] In some optional implementations of this embodiment, the step of performing cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body specifically includes: Construct a joint cost function weighted by relative velocity continuity, azimuth neighborhood overlap, and echo amplitude consistency; The minimum cost matching algorithm is used to generate cross-radar matching pairs in the same frame candidate target measurement set; Perform state estimation and update on the matching pairs, and output target trajectory parameters with trajectory labels; Unmatched measurements are marked as new trajectories or temporary trajectories.

[0035] In this embodiment, the radar data association constructs a joint cost function based on physical consistency as the criterion, and solves for the optimal correspondence through minimum cost matching. The joint cost function can be composed of a weighted sum of a relative velocity continuity term, an azimuth neighborhood overlap term, and an echo amplitude consistency term. The weights can be determined based on scene experience or offline training to ensure that kinematically reasonable candidate pairs obtain lower costs. Cross-radar matching pairs are generated on the same frame candidate target measurement set using the Hungarian algorithm or an equivalent optimization method. State estimation and updating are performed on the matching pairs. The state can include components such as position, velocity, and azimuth. α-β filtering or Kalman filtering can be used to suppress measurement noise and smooth the trajectory. After updating, target trajectory parameters with trajectory labels are output to ensure cross-frame traceability. Unmatched measurements are marked as new trajectories or temporary trajectories according to a lifetime management strategy, pending confirmation or elimination by subsequent observations to maintain the order and robustness of the trajectory library.

[0036] This application constructs a joint cost function based on relative velocity continuity, azimuth neighborhood overlap, and echo amplitude consistency, and combines minimum cost matching with state updates. This enables the uniqueness and continuity of the trajectory even when targets are dense or observations are uneven, thereby improving the accuracy of position and velocity estimation and providing core support for continuous tracking within a 180° range.

[0037] In some optional implementations of this embodiment, in the step of performing cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body, the occlusion recovery step specifically includes: If the rear millimeter-wave radar fails to generate an effective measurement of the existing trajectory in consecutive preset frames or its echo amplitude is lower than the obstruction criterion threshold, motion prediction is performed based on the trajectory parameters of the previous moment to obtain the predicted position. Search for targets matching the predicted position in the candidate measurements of the front millimeter-wave radar and perform trajectory completion. When the completion is successful, update the trajectory parameters and keep the trajectory identifier continuous. When the completion fails, set the trajectory to the observation state until effective observation is obtained or the termination threshold is reached. The updated trajectory is subjected to duplicate target merging and angle conflict resolution, and the output is limited to the target trajectory sequence within a 180° range behind the vehicle body.

[0038] In this embodiment, the occlusion recovery is initiated when the rear millimeter-wave radar detects that there are no effective measurements of the existing trajectory for several consecutive frames or the echo amplitude is lower than the occlusion criterion. The recovery process uses a motion model to extrapolate the trajectory state at the previous moment to obtain the predicted position. The model can be a uniform speed or a uniform speed turning model, and the prediction step size is consistent with the frame period to reduce time extrapolation error. Subsequently, in the candidate measurements of the front millimeter-wave radar, a matching target is searched based on the criterion of minimizing the gating radius and joint cost. If a match is successful, the trajectory is supplemented with the measurement and the trajectory identification is kept continuous. If a match fails, it enters the observation state and a termination threshold is set. If the threshold is exceeded, it is automatically cleared to prevent false trajectories from occupying the track for a long time. To avoid duplicate display and angle conflicts, the updated trajectory is subjected to overlap merging and priority arbitration, retaining only reliable trajectories, and the output is strictly limited to a range of 180 degrees behind the vehicle body, thereby providing a continuous, ghost-free backward trajectory sequence.

[0039] This application determines the missing observations of the rear millimeter-wave radar and completes the supplementary points in the front millimeter-wave radar measurement using motion prediction. It can maintain the trajectory without interruption and the trajectory marker without reset when overtaking close to the side and rear, when there is short-term obstruction or echo fading, thereby improving the recall rate and avoiding the jitter phenomenon of the prompt appearing and disappearing intermittently.

[0040] In some optional implementations of this embodiment, the steps of obtaining the head posture of the smart glasses, transforming the target trajectory sequence from the vehicle coordinate system to the head coordinate system based on pre-calibrated vehicle and head extrinsic parameters, and generating superimposed display data consistent with the target orientation and labeled with distance, specifically include: Based on the preset spatial extrinsic parameters between the vehicle body and the head and the currently acquired head posture information, coordinate transformation is performed on the position and azimuth of each target trajectory to obtain the target position and direction with the head coordinate system as a reference. Based on the coordinate transformation results, construct the corresponding graphic annotation content, output the superimposed display data containing the target's true orientation and distance, and keep the target graphic's position in the field of view consistent with the target's orientation in physical space.

[0041] In this embodiment, the coordinate transformation and overlay display generation are used to stably map the target trajectory under the vehicle system to the rider's head coordinate system and convert it into a readable visual annotation. During implementation, the head posture of the smart glasses is read, and combined with pre-calibrated spatial extrinsic parameters of the vehicle and head, the position and azimuth of the target trajectory are transformed to the head coordinate system, obtaining the target direction and distance with the head as a reference. Based on this result, overlay display data is constructed, preferably using sectors to indicate the target's orientation and concentric circles to indicate the target's distance, with relative speed or approach time information added when necessary. During rendering, it is ensured that the pointing of the displayed elements in the field of view is consistent with the target's true direction in physical space; that is, when the rider turns their head, the overlay symbol stably aligns with the target direction rather than drifting on the screen.

[0042] This application transforms the target trajectory from the vehicle coordinate system to the head coordinate system based on external participation and head posture, and generates superimposed display data with distance annotations that are consistent with the real direction. This ensures that the indication in the glasses' field of view corresponds one-to-one with the physical spatial direction, significantly reducing the rider's cognitive load and improving their intuitive judgment of the direction and distance of oncoming vehicles.

[0043] In some optional implementations of this embodiment, after the step of performing coordinate transformation on the position and azimuth of each target trajectory based on preset spatial extrinsic parameters between the vehicle body and the head and the currently acquired head posture information to obtain the target position and direction with the head coordinate system as a reference, the method further includes: The azimuth angle estimate of the target trajectory is filtered and the azimuth angle fluctuation range of consecutive frames is calculated; When the fluctuation range does not exceed the stability threshold and the synchronization error between the vehicle body and the head is less than the time synchronization threshold, the target is fixed in the corresponding direction within the field of view of the smart glasses by using the line-of-sight locking method.

[0044] In this embodiment, the display stability control is used to determine and maintain gaze-locked display after coordinate transformation. After filtering and inter-frame fluctuation statistics are performed on the transformed target azimuth sequence, if the azimuth fluctuation range of consecutive frames is less than the stability threshold and the time synchronization error between the vehicle body and the head is less than the synchronization threshold, then the gaze-locked mode is entered, fixing the target in the corresponding direction within the field of view of the smart glasses to obtain a spatially consistent visualization effect. If either condition is not met, the display is temporarily switched to a degraded display mode, with directional arrows and simplified distance labels indicating the target at the edge of the field of view, while trajectory tracking is retained and recovery conditions are continuously detected. Once the stability and synchronization requirements are met again, the display automatically returns to the gaze-locked mode. The above thresholds can be given by installation calibration or road test experience, for example, the azimuth fluctuation threshold is on the order of two degrees and the synchronization threshold is on the order of milliseconds, to balance stability and responsiveness.

[0045] This application filters the converted azimuth angle and uses the azimuth angle fluctuation range and time synchronization error as stability criteria. It enables line-of-sight locking display only when the stability conditions are met, which can effectively suppress screen jitter and pointing drift caused by small attitude disturbances or synchronization residuals, thereby improving display readability and prompt credibility.

[0046] In some optional implementations of this embodiment, the steps of obtaining echo data from the rear millimeter-wave radar located at the rear of the vehicle and the front millimeter-wave radar located at the front of the vehicle, and combining this with vehicle attitude data for time synchronization to obtain the radar measurement sequence specifically include: The vehicle controls the front and rear millimeter-wave radars to periodically transmit and receive echo signals in a time-division alternating manner. Preliminary pulse compression and frequency offset estimation are performed on the echo signals respectively; The timestamp difference is calculated based on the vehicle attitude data, and the measurement alignment is completed to output a radar measurement sequence with a synchronization timestamp.

[0047] In this embodiment, the time synchronization link ensures that the measurements from the front and rear millimeter-wave radars participate in detection and fusion on a unified time base. During implementation, the vehicle controls the two radars to periodically transmit and receive echo signals in a time-division alternating manner, avoiding mutual interference and forming a natural sampling sequence. The original echoes are first pulse-compressed to improve range resolution, and then frequency offset is calculated and compensated based on pilot signals to stabilize Doppler estimation and phase information. Subsequently, the timestamp difference between the two measurements is calculated and corrected using vehicle attitude data as a reference, outputting a radar measurement sequence with a unified timestamp for the detection and correlation modules to consume frame by frame. Actual parameters can be configured according to the system latency budget, such as a working frame rate of 20-50 Hz, an attitude update rate of no less than 100 Hz, and a synchronization residual preferably controlled within a few milliseconds to ensure the time consistency of cross-radar matching and vehicle-to-head coordinate transformation, thereby supporting subsequent stable overlay display and continuous occlusion recovery.

[0048] This application achieves time alignment by alternating the operation of the front and rear millimeter-wave radars in a time-division manner, pulse compression and frequency offset compensation, and combining attitude reference. This ensures that the two measurements participate in detection and fusion under a unified time base and stable phase conditions, reduces mutual interference, and improves the stability of Doppler and angle estimation. As a result, it provides highly consistent raw input for cross-radar correlation and head coordinate transformation and meets real-time requirements.

[0049] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0050] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0051] Further reference Figure 2 As a response to the above Figure 1 The implementation of the method shown in this application provides an embodiment of a bicycle rearward perception system based on multiple radars and smart glasses. This system embodiment is similar to... Figure 1 Corresponding to the method embodiments shown, the system can be specifically applied to various electronic devices.

[0052] like Figure 2 As shown, the bicycle rearward perception system 200 based on multiple radars and smart glasses described in this embodiment includes: an acquisition module 201, an output module 202, an execution module 203, a conversion module 204, and a display module 205. Wherein: The acquisition module 201 is used to acquire the echo data of the rear millimeter-wave radar set at the rear of the vehicle body and the front millimeter-wave radar set at the front of the vehicle body, and to synchronize the data with the vehicle attitude data to obtain the radar measurement sequence. Output module 202 is used to perform target detection and parameter estimation on the radar measurement sequence, output a target measurement set including target distance and azimuth angle, and organize the target measurement set in the vehicle coordinate system; The execution module 203 is used to perform cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar, and generate a target trajectory sequence covering a 180° range behind the vehicle body. The conversion module 204 is used to obtain the head posture of the smart glasses, convert the target trajectory sequence from the vehicle coordinate system to the head coordinate system according to the pre-calibrated vehicle and head extrinsic parameters, and generate superimposed display data that is consistent with the target orientation and marked with distance; Display module 205 is used to render the overlay display data within the field of view of the smart glasses.

[0053] The bicycle rear perception system based on multiple radars and smart glasses provided in this embodiment of the invention can realize all the processes of the bicycle rear perception method based on multiple radars and smart glasses in the above embodiments. The functions and technical effects of each module in the device are the same as those of the bicycle rear perception method based on multiple radars and smart glasses in the above embodiments, and will not be repeated here.

[0054] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 3 , Figure 3 This is a basic structural block diagram of the computer device in this embodiment.

[0055] The computer device 3 includes a memory 31, a processor 32, and a network interface 33 that are interconnected via a system bus. It should be noted that only the computer device 3 with components 31-33 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0056] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0057] The memory 31 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 31 may be an internal storage unit of the computer device 3, such as the hard disk or memory of the computer device 3. In other embodiments, the memory 31 may also be an external storage device of the computer device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 3. Of course, the memory 31 may also include both the internal storage unit and its external storage device of the computer device 3. In this embodiment, the memory 31 is typically used to store the operating system and various application software installed on the computer device 3, such as computer-readable instructions for a bicycle rear-view perception method based on multi-radar and smart glasses. In addition, the memory 31 can also be used to temporarily store various types of data that have been output or will be output.

[0058] In some embodiments, the processor 32 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 32 is typically used to control the overall operation of the computer device 3. In this embodiment, the processor 32 is used to execute computer-readable instructions stored in the memory 31 or to process data, for example, to execute the computer-readable instructions of the bicycle rear-view perception method based on multi-radar and smart glasses.

[0059] The network interface 33 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 3 and other electronic devices.

[0060] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the bicycle rear perception method based on multiple radars and smart glasses as described above.

[0061] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0062] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A bicycle rearward perception method based on multiple radars and smart glasses, characterized in that, Includes the following steps: The echo data of the rear millimeter-wave radar located at the rear of the vehicle body and the front millimeter-wave radar located at the front of the vehicle body are acquired and time-synchronized with the vehicle attitude data to obtain the radar measurement sequence. Target detection and parameter estimation are performed on the radar measurement sequence, and a target measurement set containing target distance and azimuth is output. The target measurement set is then organized in the vehicle coordinate system. Perform cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body; The head posture of the smart glasses is obtained, and the target trajectory sequence is transformed from the vehicle coordinate system to the head coordinate system based on the pre-calibrated vehicle and head extrinsic parameters to generate superimposed display data that is consistent with the target orientation and marked with distance. The overlay display data is rendered within the field of view of the smart glasses.

2. The bicycle rearward perception method based on multiple radars and smart glasses according to claim 1, characterized in that, Before the step of performing cross-radar data correlation and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle, the method further includes: The target measurement data from the front and rear millimeter-wave radars are unified into the vehicle coordinate system and time-aligned according to the synchronization timestamp. Calculate the distance difference, azimuth difference, and relative velocity difference separately, and set multidimensional gating thresholds to eliminate measurements that exceed the thresholds; Preserve the same frame candidate target measurement set used for cross-radar data correlation.

3. The bicycle rearward perception method based on multiple radars and smart glasses according to claim 2, characterized in that, In the step of performing cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body, the radar data association step specifically includes: Construct a joint cost function weighted by relative velocity continuity, azimuth neighborhood overlap, and echo amplitude consistency; The minimum cost matching algorithm is used to generate cross-radar matching pairs in the same frame candidate target measurement set; Perform state estimation and update on the matching pairs, and output target trajectory parameters with trajectory labels; Unmatched measurements are marked as new trajectories or temporary trajectories.

4. The bicycle rearward perception method based on multiple radars and smart glasses according to claim 3, characterized in that, In the step of performing cross-radar data correlation and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar to generate a target trajectory sequence covering a 180° range behind the vehicle body, the occlusion recovery step specifically includes: If the rear millimeter-wave radar fails to generate an effective measurement of the existing trajectory in consecutive preset frames or its echo amplitude is lower than the obstruction criterion threshold, motion prediction is performed based on the trajectory parameters of the previous moment to obtain the predicted position. Search for targets matching the predicted position in the candidate measurements of the front millimeter-wave radar and perform trajectory completion. When the completion is successful, update the trajectory parameters and keep the trajectory identifier continuous. When the completion fails, set the trajectory to the observation state until effective observation is obtained or the termination threshold is reached. The updated trajectory is subjected to duplicate target merging and angle conflict resolution, and the output is limited to the target trajectory sequence within a 180° range behind the vehicle body.

5. The bicycle rearward perception method based on multiple radars and smart glasses according to claim 1, characterized in that, The steps of acquiring the head posture of the smart glasses, transforming the target trajectory sequence from the vehicle coordinate system to the head coordinate system based on pre-calibrated vehicle and head extrinsic parameters, and generating superimposed display data consistent with the target orientation and labeled with distance, specifically include: Based on the preset spatial extrinsic parameters between the vehicle body and the head and the currently acquired head posture information, coordinate transformation is performed on the position and azimuth of each target trajectory to obtain the target position and direction with the head coordinate system as a reference. Based on the coordinate transformation results, construct the corresponding graphic annotation content, output the superimposed display data containing the target's true orientation and distance, and keep the target graphic's position in the field of view consistent with the target's orientation in physical space.

6. The bicycle rearward perception method based on multiple radars and smart glasses according to claim 5, characterized in that, After the step of performing coordinate transformation on the position and azimuth of each target trajectory based on preset spatial extrinsic parameters between the vehicle body and the head and the currently acquired head posture information to obtain the target position and orientation with the head coordinate system as a reference, the method further includes: The azimuth angle estimate of the target trajectory is filtered and the azimuth angle fluctuation range of consecutive frames is calculated; When the fluctuation range does not exceed the stability threshold and the synchronization error between the vehicle body and the head is less than the time synchronization threshold, the target is fixed in the corresponding direction within the field of view of the smart glasses by using the line-of-sight locking method.

7. The bicycle rearward perception method based on multiple radars and smart glasses according to claim 1, characterized in that, The step of acquiring echo data from the rear millimeter-wave radar located at the rear of the vehicle and the front millimeter-wave radar located at the front of the vehicle, and combining this with vehicle attitude data for time synchronization to obtain the radar measurement sequence specifically includes: The vehicle controls the front and rear millimeter-wave radars to periodically transmit and receive echo signals in a time-division alternating manner. Preliminary pulse compression and frequency offset estimation are performed on the echo signals respectively; The timestamp difference is calculated based on the vehicle attitude data, and the measurement alignment is completed to output a radar measurement sequence with a synchronization timestamp.

8. A bicycle rearward perception system based on multiple radars and smart glasses, characterized in that, include: The acquisition module is used to acquire the echo data of the rear millimeter-wave radar located at the rear of the vehicle body and the front millimeter-wave radar located at the front of the vehicle body, and to combine the vehicle attitude data for time synchronization to obtain the radar measurement sequence. The output module is used to perform target detection and parameter estimation on the radar measurement sequence, output a target measurement set including target distance and azimuth, and organize the target measurement set in the vehicle coordinate system; The execution module is used to perform cross-radar data association and occlusion recovery on the target measurement sets from the front millimeter-wave radar and the rear millimeter-wave radar, and generate a target trajectory sequence covering a 180° range behind the vehicle body; The conversion module is used to obtain the head posture of the smart glasses, and convert the target trajectory sequence from the vehicle coordinate system to the head coordinate system according to the pre-calibrated vehicle and head extrinsic parameters, and generate superimposed display data that is consistent with the target orientation and marked with distance; The display module is used to render the overlay display data within the field of view of the smart glasses.

9. A computer device, characterized in that, The method includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the bicycle rearward perception method based on multiple radars and smart glasses as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the bicycle rearward perception method based on multiple radars and smart glasses as described in any one of claims 1 to 7.

Images