Methods for outdoor sports safety early warning and trajectory tracking protection for hearing-impaired individuals

Abstract

This invention provides a method for outdoor sports safety early warning and trajectory tracking protection for hearing-impaired individuals, relating to the field of safety technology. The method includes acquiring the location of the hearing-impaired individual and dynamic environmental factors, assessing collision risk through a trajectory prediction algorithm, selecting and adapting non-acoustic sensing signals based on risk levels, and sending dynamically adjusted directional instructions to a wearable device. This invention can identify hazards in real time, accurately predict collision risks, and present personalized non-acoustic warning signals, effectively improving the safety of hearing-impaired individuals during outdoor activities and reducing the incidence of accidents.

Description

Technical Field

[0001] This invention relates to the field of safety technology, and in particular to a method for safety early warning and trajectory tracking protection for hearing-impaired individuals during outdoor exercise. Background Technology

[0002] As society shows increasing care for people with disabilities, the social integration and quality of life of the hearing-impaired are receiving growing attention. Due to their impaired auditory perception, the hearing-impaired often face safety risks when outdoors, especially in complex urban environments, where they cannot perceive danger signals from their surroundings, such as vehicle horns, construction warnings, or other dynamically approaching hazards. Currently, assistive technologies for the hearing-impaired primarily focus on indoor environments or communication assistance, while systematic solutions for ensuring safety during outdoor activities remain insufficient.

[0003] Existing early warning systems lack targeted design for the specific needs of hearing-impaired individuals. Most safety early warning devices rely primarily on sound prompts, which cannot effectively provide timely and effective danger warnings for hearing-impaired users. Furthermore, a few specialized devices often use simple vibration or light signals, which cannot convey complex danger information and directional indications.

[0004] Existing technologies lack the ability to intelligently identify and classify dynamic environmental elements, cannot distinguish between static obstacles and dynamic approach elements with collision risks, and cannot formulate corresponding safety response strategies based on the motion pattern characteristics of different dynamic elements, resulting in insufficient accuracy of early warnings and a tendency to produce false alarms or missed alarms. Summary of the Invention

[0005] This invention provides a method for outdoor sports safety early warning and trajectory tracking protection for hearing-impaired individuals, which can solve the problems in the prior art.

[0006] A first aspect of the present invention provides a method for outdoor sports safety early warning and trajectory tracking protection for hearing-impaired individuals, comprising:

[0007] The system acquires real-time location and motion status information of the target hearing-impaired person in an outdoor environment, continuously monitors changes in dynamic elements of the environment in which the target hearing-impaired person is located, classifies and labels the changes in dynamic elements, distinguishes between static obstacle elements and dynamic approach elements, and extracts the motion pattern features of the dynamic approach elements.

[0008] Based on the motion state information and the motion pattern characteristics of the dynamic approach elements, a trajectory prediction algorithm is used to generate the future position distribution of the target hearing-impaired person and the dynamic approach elements. The minimum safe distance between the position distributions is calculated. When the minimum safe distance is lower than the safety threshold corresponding to the motion pattern characteristics, a collision risk is determined and a risk response level is identified.

[0009] Based on the motion pattern characteristics and the risk response level, and combined with a pre-built multi-level signal type screening rule library, non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person are screened, and the presentation sequence of the non-acoustic perception signals is determined according to the current motion state of the target hearing-impaired person.

[0010] Based on the non-acoustic sensing signal, a sensing command is sent to the wearable interactive device of the target hearing-impaired person, and the directional indication information of the non-acoustic sensing signal is dynamically adjusted according to the real-time position change of the dynamic proximity element to guide the target hearing-impaired person to adjust their movement trajectory.

[0011] The dynamic element changes are classified and labeled to distinguish between static obstacle elements and dynamic approach elements, and the motion pattern features of the dynamic approach elements are extracted, including:

[0012] The dynamic element changes are continuously identified. By analyzing the magnitude and frequency of element position changes within a preset observation period, elements with position changes below a position stability threshold are labeled as static obstacle elements, while elements with position changes exceeding the position stability threshold and whose distance from the target hearing-impaired person shows a decreasing trend are labeled as dynamic approach elements.

[0013] For objects labeled as dynamic proximity elements, a motion trajectory sequence is established, and pattern clustering analysis is performed on the motion trajectory sequence to identify the dominant motion direction and motion periodicity characteristics of the dynamic proximity elements; based on the curvature change characteristics and velocity fluctuation characteristics of the motion trajectory sequence, the motion stability level of the dynamic proximity elements is determined.

[0014] The motion stability level, the dominant motion direction, and the motion periodicity characteristics together constitute the motion pattern characteristics.

[0015] Pattern clustering analysis is performed on the motion trajectory sequence to identify the dominant motion direction and periodicity characteristics of the dynamic approach element; based on the curvature change characteristics and velocity fluctuation characteristics of the motion trajectory sequence, the motion stability level of the dynamic approach element is determined, including:

[0016] The direction vectors of the trajectory points in the motion trajectory sequence are extracted, the direction vectors are projected onto a preset direction division space, the distribution density of trajectory points in each direction interval is calculated, and the direction interval with the highest distribution density is taken as the dominant motion direction.

[0017] Calculate the time interval and displacement change between adjacent trajectory points in the motion trajectory sequence to construct a time-displacement correlation sequence. Perform periodic detection on the time-displacement correlation sequence, extract its repetition period, and use the repetition period as a motion periodic feature.

[0018] The curvature value is calculated for the trajectory segment formed by three consecutive trajectory points in the motion trajectory sequence, a curvature change sequence is generated, the variance and extreme value distribution of the curvature change sequence are statistically analyzed, and the variance and extreme value distribution are normalized and then weighted and summed to determine the trajectory smoothness;

[0019] The instantaneous velocities corresponding to each trajectory point in the motion trajectory sequence are sampled to generate a velocity change sequence. The standard deviation and coefficient of variation of the velocity change sequence are calculated. After normalizing the standard deviation and the coefficient of variation, a weighted average is performed to determine the velocity stability.

[0020] When both the trajectory smoothness and the velocity stability exceed the corresponding judgment threshold, the motion stability level is determined to be a high stability level; when either the trajectory smoothness or the velocity stability is less than the corresponding judgment threshold, the motion stability level is determined to be a low stability level.

[0021] Based on the motion state information and the motion pattern characteristics of the dynamic proximity element, a trajectory prediction algorithm is used to generate the future position distribution of the target hearing-impaired person and the dynamic proximity element, and the minimum safe distance between the position distributions is calculated, including:

[0022] Based on the current position coordinates, current velocity vector, and current direction of motion corresponding to the motion state information, and combined with the motion inertial parameters of the target hearing-impaired person, the trajectory evolution process of the target hearing-impaired person in the future predicted time period is deduced, and a set of first predicted position points of the target hearing-impaired person at multiple future moments is generated;

[0023] The baseline motion trend of the dynamic approach element is determined based on the dominant motion direction. The periodic offset in the baseline motion trend is corrected based on the periodicity characteristics of the motion. The uncertainty range of trajectory prediction is determined based on the motion stability level. Based on the corrected motion trend and the uncertainty range, a second set of predicted position points of the dynamic approach element corresponding to the multiple future moments is generated.

[0024] Calculate the spatial distance between each predicted location point in the first predicted location point set and the second predicted location point set at the same time. Iterate through all times within the future prediction time period and extract the minimum value of the spatial distance as the minimum safe distance between the location distributions.

[0025] Based on the motion pattern characteristics and the risk response level, and combined with a pre-built multi-level signal type screening rule base, non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person are screened. Furthermore, based on the current motion state of the target hearing-impaired person, the presentation sequence of the non-acoustic perception signals is determined, including:

[0026] A multi-level signal type filtering rule base is established, which includes a first-level filtering rule for risk response level, a second-level filtering rule for motion stability level, and a third-level filtering rule for dominant motion direction.

[0027] The initial candidate signal type set contains all available non-acoustic sensing signal types. The motion stability level and dominant motion direction of the dynamic proximity element are extracted from the motion pattern features. The risk response level is substituted into the first layer of screening rules. Signal types that do not meet the risk response level requirements are excluded from the initial candidate signal type set to obtain the first screening result set.

[0028] Substitute the motion stability level into the second layer of filtering rules, and exclude signal types that do not meet the motion stability level requirements from the first filtering result set to obtain the second filtering result set;

[0029] Substitute the dominant motion direction into the third-level filtering rules, and select the signal type that meets the requirement of the dominant motion direction from the second filtering result set as the finally determined non-acoustic sensing signal type;

[0030] Extract the current movement speed of the target hearing-impaired person from the current movement state, determine the reaction time margin of the target hearing-impaired person based on the current movement speed, calculate the danger signal trigger time based on the reaction time margin and the minimum safe distance, and use the danger signal trigger time as the starting point of the presentation sequence of the non-acoustic sensing signal.

[0031] Based on the real-time positional changes of the dynamic proximity elements, the directional indication information of the non-acoustic sensing signals is dynamically adjusted to guide the target hearing-impaired person to adjust their movement trajectory, including:

[0032] A spatial direction reference coordinate system is established with the current position of the target hearing-impaired person as the origin. The displacement vector between the current position of the dynamic approach element and the previous position of the dynamic approach element is calculated. The approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person is determined based on the displacement vector. The approach azimuth angle is mapped to the spatial direction reference coordinate system to obtain the direction of the hazard source.

[0033] An initial avoidance direction sector is determined based on the direction of the hazard source, wherein the initial avoidance direction sector is a spatial region opposite to the direction of the hazard source.

[0034] Obstacle distribution information of the surrounding environment of the target hearing-impaired person is obtained. The positions of static obstacles within the fan-shaped area of ​​the initial avoidance direction are identified from the obstacle distribution information. Sub-regions in the fan-shaped area of ​​the initial avoidance direction that do not overlap with the positions of the static obstacles are marked as feasible avoidance areas. The center direction vector of the feasible avoidance area is calculated as the preferred avoidance direction.

[0035] Directional indication information is generated based on the preferred avoidance direction. The directional indication information includes the turning angle and turning side relative to the current orientation of the target hearing-impaired person. The directional indication information is encoded into the non-acoustic perception signal. The non-acoustic perception signal containing the directional indication information is output to the target hearing-impaired person through the wearable interactive device to guide the target hearing-impaired person to adjust their movement trajectory.

[0036] The approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person is determined based on the displacement vector. The approach azimuth angle is then mapped to the spatial direction reference coordinate system to obtain the direction of the hazard source, including:

[0037] Obtain the current orientation vector of the target hearing-impaired person, the current orientation vector representing the direction the target hearing-impaired person's body is pointing forward, calculate the vector angle between the displacement vector and the current orientation vector to obtain the relative azimuth angle of the displacement vector relative to the current orientation vector;

[0038] Determine the quadrant interval in which the relative azimuth angle is located, convert the relative azimuth angle into a somatosensory azimuth angle with reference to the body of the target hearing-impaired person based on the quadrant interval, and determine the somatosensory azimuth angle as the approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person;

[0039] An azimuth mapping rule table is established, which stores the mapping relationship between each region identifier of the perceived azimuth and the corresponding direction interval in the spatial direction reference coordinate system. The azimuth mapping rule table is queried according to the region identifier close to the azimuth, and the direction interval of the spatial direction reference coordinate system corresponding to the region identifier is obtained. The center direction angle of the direction interval of the spatial direction reference coordinate system is calculated, and the direction vector corresponding to the center direction angle is determined as the direction of the hazard source.

[0040] A second aspect of the present invention provides a safety early warning and trajectory tracking protection system for outdoor sports activities of hearing-impaired individuals, comprising:

[0041] The first unit is used to acquire real-time location and motion status information of the target hearing-impaired person in an outdoor environment, continuously monitor the dynamic changes of the environment in which the target hearing-impaired person is located, classify and label the dynamic changes, distinguish between static obstacle elements and dynamic approach elements, and extract the motion pattern features of the dynamic approach elements;

[0042] The second unit is used to generate the future position distribution of the target hearing-impaired person and the dynamic approach element based on the motion state information and the motion pattern characteristics of the dynamic approach element through a trajectory prediction algorithm, calculate the minimum safe distance between the position distributions, and determine the risk of collision and the risk response level when the minimum safe distance is lower than the safety threshold corresponding to the motion pattern characteristics.

[0043] The third unit is used to filter non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person based on the motion pattern characteristics and the risk response level, combined with a pre-built multi-level signal type filtering rule library, and to determine the presentation sequence of the non-acoustic perception signals based on the current motion state of the target hearing-impaired person;

[0044] The fourth unit is used to send sensing commands to the wearable interactive device of the target hearing-impaired person based on the non-acoustic sensing signals, and dynamically adjust the direction indication information of the non-acoustic sensing signals according to the real-time position changes of the dynamic proximity elements, so as to guide the target hearing-impaired person to adjust the movement trajectory.

[0045] A third aspect of the embodiments of the present invention,

[0046] An electronic device is provided, comprising:

[0047] processor;

[0048] Memory used to store processor-executable instructions;

[0049] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0050] Fourth aspect of the present invention,

[0051] A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0052] The beneficial effects of this application are as follows:

[0053] By monitoring the dynamic changes in the surrounding environment of hearing-impaired individuals in real time, intelligent classification and assessment of different types of potential risks are performed. Combined with trajectory prediction algorithms, the minimum safe distance is accurately calculated, which can identify potential collision risks in advance and significantly improve the safety of hearing-impaired individuals in outdoor environments.

[0054] Based on a multi-level signal type filtering rule base, this system intelligently selects the most suitable non-acoustic perception signal type according to the individual perception characteristics of hearing-impaired individuals, and dynamically adjusts the signal presentation sequence according to the motion state. This solves the problem that traditional acoustic warnings are ineffective for hearing-impaired individuals, and achieves accuracy and timeliness in information transmission.

[0055] By providing dynamically adjustable directional guidance information through wearable interactive devices, hearing-impaired individuals can adjust their movement trajectories in a timely manner to avoid potential risks, while maintaining their autonomy in movement. This ensures the effectiveness of safety protection while respecting the independent action capabilities of hearing-impaired individuals, thereby improving their safety experience and quality of life during outdoor activities. Attached Figure Description

[0056] Figure 1 This is a flowchart illustrating the method for outdoor sports safety early warning and trajectory tracking protection for hearing-impaired individuals, as described in an embodiment of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0059] Figure 1 This is a flowchart illustrating an embodiment of the present invention regarding a method for outdoor sports safety early warning and trajectory tracking protection for hearing-impaired individuals. Figure 1 As shown, the method includes:

[0060] The system acquires real-time location and motion status information of the target hearing-impaired person in an outdoor environment, continuously monitors changes in dynamic elements of the environment in which the target hearing-impaired person is located, classifies and labels the changes in dynamic elements, distinguishes between static obstacle elements and dynamic approach elements, and extracts the motion pattern features of the dynamic approach elements.

[0061] Based on the motion state information and the motion pattern characteristics of the dynamic approach elements, a trajectory prediction algorithm is used to generate the future position distribution of the target hearing-impaired person and the dynamic approach elements. The minimum safe distance between the position distributions is calculated. When the minimum safe distance is lower than the safety threshold corresponding to the motion pattern characteristics, a collision risk is determined and a risk response level is identified.

[0062] Based on the motion pattern characteristics and the risk response level, and combined with a pre-built multi-level signal type screening rule library, non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person are screened, and the presentation sequence of the non-acoustic perception signals is determined according to the current motion state of the target hearing-impaired person.

[0063] Based on the non-acoustic sensing signal, a sensing command is sent to the wearable interactive device of the target hearing-impaired person, and the directional indication information of the non-acoustic sensing signal is dynamically adjusted according to the real-time position change of the dynamic proximity element to guide the target hearing-impaired person to adjust their movement trajectory.

[0064] In one optional implementation, the changes in the dynamic elements are classified and labeled to distinguish between static obstacle elements and dynamic approach elements, and the motion pattern features of the dynamic approach elements are extracted, including:

[0065] The dynamic element changes are continuously identified. By analyzing the magnitude and frequency of element position changes within a preset observation period, elements with position changes below a position stability threshold are labeled as static obstacle elements, while elements with position changes exceeding the position stability threshold and whose distance from the target hearing-impaired person shows a decreasing trend are labeled as dynamic approach elements.

[0066] For objects labeled as dynamic proximity elements, a motion trajectory sequence is established, and pattern clustering analysis is performed on the motion trajectory sequence to identify the dominant motion direction and motion periodicity characteristics of the dynamic proximity elements; based on the curvature change characteristics and velocity fluctuation characteristics of the motion trajectory sequence, the motion stability level of the dynamic proximity elements is determined.

[0067] The motion stability level, the dominant motion direction, and the motion periodicity characteristics together constitute the motion pattern characteristics.

[0068] In implementing this invention, dynamic element changes are first classified and labeled, and environmental elements are distinguished into static obstacle elements and dynamic approach elements. The motion pattern features of dynamic approach elements are extracted in order to provide more accurate safety prompts for the target hearing-impaired person.

[0069] The determination of the persistence of dynamic element changes is achieved by analyzing the changes in element position within a preset observation period. Specifically, multi-frame image capture technology is used, acquiring one frame of environmental image every 100 milliseconds, with 30 consecutive frames constituting one observation period. For each detected element, the difference in its positional coordinates between adjacent frames is calculated, and the magnitude of the element's positional change within the observation period is accumulated. When the magnitude of the element's positional change is lower than the positional stability threshold, it is marked as a static obstacle element. Here, the positional stability threshold is set to 0.5 meters, meaning that if the element's positional change does not exceed 0.5 meters within a 3-second observation period, it is considered a static obstacle element. Conversely, when the magnitude of the element's positional change exceeds 0.5 meters, and the distance between it and the target hearing-impaired person shows a decreasing trend, it is marked as a dynamic approaching element. The distance decreasing trend is determined by calculating the rate of change of distance between the element and the hearing-impaired person over 5 consecutive frames. If the average rate of change is negative and the absolute value is greater than 0.2 meters per second, it is confirmed as a distance decreasing trend.

[0070] For objects labeled as dynamic proximity features, a motion trajectory sequence is established by recording the spatial coordinates of the feature over multiple consecutive observation periods. The position coordinates of the dynamic proximity feature are recorded over 10 consecutive observation periods (a total of 30 seconds), forming a motion trajectory sequence containing 300 spatiotemporal points. Based on this trajectory data, pattern clustering analysis is performed to identify the dominant motion direction and periodicity characteristics of the dynamic proximity feature.

[0071] The dominant direction of motion is identified by calculating the direction vectors between adjacent points in the trajectory sequence and then performing statistical analysis on these vectors. The 360-degree direction is divided into eight 45-degree intervals, and the number of vectors falling into each interval is counted. When the number of vectors in a particular interval exceeds 60%, that direction is determined to be the dominant direction of motion. For example, if the northward (337.5 degrees to 22.5 degrees) motion vectors account for 65%, then northward is identified as the dominant direction of motion for this dynamic approach element.

[0072] The identification of periodic motion characteristics is achieved by analyzing the patterns of velocity and direction changes in the trajectory sequence. The frequency of peaks and troughs in the velocity values ​​of the trajectory sequence is calculated; when a regular velocity fluctuation pattern is detected, it is determined to be periodic motion. A velocity fluctuation threshold of 0.5 m / s is set. When the velocity fluctuation amplitude exceeds this threshold and the fluctuation interval is relatively stable (standard deviation of the fluctuation interval is less than 0.3 seconds), the period value is recorded. For example, if the velocity of a certain dynamic approach element shows a significant fluctuation every 2.1 seconds with a standard deviation of 0.25 seconds, then 2.1 seconds is determined as its motion period.

[0073] The motion stability level of a dynamic approach element is determined based on the curvature variation characteristics and velocity fluctuation characteristics of the motion trajectory sequence. Curvature variation characteristics are characterized by calculating the curvature value formed by every three consecutive points in the trajectory sequence, and then calculating the average curvature and standard deviation of the entire trajectory. Velocity fluctuation characteristics are characterized by calculating the standard deviation of the velocity values ​​in the trajectory sequence. Motion stability levels are divided into high, medium, and low. A high stability level is defined as an average curvature less than 0.2, a standard deviation of curvature less than 0.1, and a standard deviation of velocity less than 0.3 m / s. A medium stability level is defined as an average curvature between 0.2 and 0.5, a standard deviation of curvature between 0.1 and 0.3, or a standard deviation of velocity between 0.3 and 0.8 m / s. All other cases are defined as low stability.

[0074] For example, the trajectory data of an electric bicycle shows an average curvature of 0.15, a standard deviation of curvature of 0.08, and a speed standard deviation of 0.25 m / s, classifying its stability level as high. In contrast, the trajectory data of a running pedestrian shows an average curvature of 0.35, a standard deviation of curvature of 0.22, and a speed standard deviation of 0.65 m / s, classifying its stability level as medium.

[0075] Ultimately, the dominant direction of movement, the periodicity of movement, and the stability level are combined to form the movement pattern characteristics of dynamic approach elements. For example, the movement pattern characteristics of a certain dynamic approach element can be represented as: dominant direction is northeast, movement period is irregular, and stability level is medium. Another example is: dominant direction is south, movement period is 1.8 seconds, and stability level is high. These movement pattern characteristics provide important basic information for the subsequent generation of auditory cues, enabling the generation of differentiated auditory warnings for dynamic approach elements with different characteristics, thereby improving the hearing-impaired person's perception of their surroundings and their level of safety.

[0076] In one optional implementation, pattern clustering analysis is performed on the motion trajectory sequence to identify the dominant motion direction and periodicity characteristics of the dynamic approach element; based on the curvature change characteristics and velocity fluctuation characteristics of the motion trajectory sequence, the motion stability level of the dynamic approach element is determined, including:

[0077] The direction vectors of the trajectory points in the motion trajectory sequence are extracted, the direction vectors are projected onto a preset direction division space, the distribution density of trajectory points in each direction interval is calculated, and the direction interval with the highest distribution density is taken as the dominant motion direction.

[0078] Calculate the time interval and displacement change between adjacent trajectory points in the motion trajectory sequence to construct a time-displacement correlation sequence. Perform periodic detection on the time-displacement correlation sequence, extract its repetition period, and use the repetition period as a motion periodic feature.

[0079] The curvature value is calculated for the trajectory segment formed by three consecutive trajectory points in the motion trajectory sequence, a curvature change sequence is generated, the variance and extreme value distribution of the curvature change sequence are statistically analyzed, and the variance and extreme value distribution are normalized and then weighted and summed to determine the trajectory smoothness;

[0080] The instantaneous velocities corresponding to each trajectory point in the motion trajectory sequence are sampled to generate a velocity change sequence. The standard deviation and coefficient of variation of the velocity change sequence are calculated. After normalizing the standard deviation and the coefficient of variation, a weighted average is performed to determine the velocity stability.

[0081] When both the trajectory smoothness and the velocity stability exceed the corresponding judgment threshold, the motion stability level is determined to be a high stability level; when either the trajectory smoothness or the velocity stability is less than the corresponding judgment threshold, the motion stability level is determined to be a low stability level.

[0082] When performing pattern clustering analysis on a motion trajectory sequence, the first step is to extract the direction vectors of the trajectory points in the sequence. Specifically, for each trajectory point P in the motion trajectory sequence... i Calculate its relationship with the next trajectory point P. i+1 The direction of the line connecting them forms the direction vector V. i For example, for trajectory points P1(10, 15) and P2(12, 20), the calculated direction vector V1 is (2, 5), representing the direction of movement from P1 to P2. The obtained direction vector is projected onto a preset direction division space, which can be set to eight direction intervals: North (337.5°-22.5°), Northeast (22.5°-67.5°), East (67.5°-112.5°), Southeast (112.5°-157.5°), South (157.5°-202.5°), Southwest (202.5°-247.5°), West (247.5°-292.5°), and Northwest (292.5°-337.5°). The distribution density of trajectory points in each directional interval is statistically analyzed. For example, in a sequence containing 100 trajectory points, there are 45 points in the north interval, 20 points in the northeast interval, and 10, 8, 7, 5, 3 and 2 points in other intervals, respectively. The distribution density in the north interval is the highest, and the north is determined as the dominant movement direction of this dynamic proximity element.

[0083] To extract the periodicity features of motion, the time interval and displacement change between adjacent trajectory points in the motion trajectory sequence are calculated to construct a time-displacement correlation sequence. For example, for three trajectory points with timestamps t1=1.0s, t2=1.5s, and t3=2.0s, and positions P1(10,15), P2(12,20), and P3(14,25), the calculated time intervals are Δt1=0.5s and Δt2=0.5s, and the displacement changes are Δd1=5.4m and Δd2=5.8m, respectively, thus constructing a time-displacement correlation sequence {(0.5s, 5.4m), (0.5s, 5.8m)}. Periodicity detection of this time-displacement correlation sequence can be performed using autocorrelation analysis to identify repeating patterns in the sequence. If a similar displacement pattern is detected every 2.5 seconds, the motion period is determined to be 2.5 seconds. This period value serves as the periodicity feature of the dynamic proximity element.

[0084] To analyze the curvature variation characteristics, the curvature value is calculated for the trajectory segment formed by three consecutive trajectory points in the motion trajectory sequence, generating a curvature variation sequence. Specifically, for three consecutive trajectory points P1(10,15), P2(12,20), and P3(14,25), the curvature value K=1 / R is obtained by calculating the reciprocal of the radius R of the circle determined by these three points. The curvature values ​​of all three consecutive points in the trajectory sequence are calculated to obtain the curvature variation sequence {K1, K2, ..., Kn}. The variance and extreme value distribution of this curvature variation sequence are statistically analyzed. For example, the variance is 0.015, the maximum curvature value is 0.08, and the minimum curvature value is 0.01. The variance and extreme value distribution are normalized. The normalized variance can be set to 0.6, and the extreme value distribution characteristic value can be set to 0.7. Weighted summation is performed on the normalized variance and extreme value distribution characteristic value using weighting coefficients of 0.6 and 0.4, respectively, to obtain the trajectory smoothness.

[0085] To analyze the velocity fluctuation characteristics, the instantaneous velocities corresponding to each trajectory point in the motion trajectory sequence are sampled to generate a velocity change sequence. Instantaneous velocities can be calculated by removing the positions of adjacent trajectory points and dividing by the time interval. For example, the instantaneous velocity from trajectory point P1 to P2 is v1 = 5.4 m / 0.5 s = 10.8 m / s, and the instantaneous velocity from trajectory point P2 to P3 is v2 = 5.8 m / s / 0.5 s = 11.6 m / s, thus obtaining the velocity change sequence {10.8 m / s, 11.6 m / s, ...}. The standard deviation and coefficient of variation of this velocity change sequence are calculated; for example, the standard deviation is 1.2 m / s, and the coefficient of variation (standard deviation divided by the average velocity) is 0.11. The standard deviation and coefficient of variation are normalized. The normalized standard deviation can be set to 0.3 and the normalized coefficient of variation to 0.25. The normalized standard deviation and coefficient of variation are weighted and averaged using weighting coefficients of 0.5 and 0.5 respectively, resulting in a speed stability of 0.275.

[0086] Finally, the motion stability level is determined based on the calculated trajectory smoothness and velocity stability. Judgment thresholds are set: 0.6 for trajectory smoothness and 0.3 for velocity stability. When the trajectory smoothness (0.64) is greater than the judgment threshold of 0.6, and the velocity stability (0.275) is less than the judgment threshold of 0.3, the motion stability level of the dynamic approach feature is determined to be low stability. If both trajectory smoothness and velocity stability exceed their respective judgment thresholds, such as trajectory smoothness of 0.7 and velocity stability of 0.4, the motion stability level is determined to be high stability. In this way, the motion state of the dynamic approach feature can be comprehensively evaluated, providing a basis for subsequent risk assessment and decision-making.

[0087] In one optional implementation, based on the motion state information and the motion pattern characteristics of the dynamic proximity element, a trajectory prediction algorithm is used to generate the future position distribution of the target hearing-impaired person and the dynamic proximity element, and the minimum safe distance between the position distributions is calculated, including:

[0088] Based on the current position coordinates, current velocity vector, and current direction of motion corresponding to the motion state information, and combined with the motion inertial parameters of the target hearing-impaired person, the trajectory evolution process of the target hearing-impaired person in the future predicted time period is deduced, and a set of first predicted position points of the target hearing-impaired person at multiple future moments is generated;

[0089] The baseline motion trend of the dynamic approach element is determined based on the dominant motion direction. The periodic offset in the baseline motion trend is corrected based on the periodicity characteristics of the motion. The uncertainty range of trajectory prediction is determined based on the motion stability level. Based on the corrected motion trend and the uncertainty range, a second set of predicted position points of the dynamic approach element corresponding to the multiple future moments is generated.

[0090] Calculate the spatial distance between each predicted location point in the first predicted location point set and the second predicted location point set at the same time. Iterate through all times within the future prediction time period and extract the minimum value of the spatial distance as the minimum safe distance between the location distributions.

[0091] In this embodiment, a method for generating trajectory prediction based on motion state information and motion pattern features of dynamic proximity elements is provided. This method is used to predict the positional distribution of the target hearing-impaired person and dynamic proximity elements at future times and to calculate the minimum safe distance.

[0092] The trajectory prediction algorithm first obtains the motion state information of the target hearing-impaired person, including their current position coordinates (x0, y0), current velocity vector (vx, vy), and current direction of motion θ. It also needs to obtain the target hearing-impaired person's motion inertial parameters, such as reaction time tr (0.8 seconds) and rate of change of acceleration a (0.5 m / s²). 2 The rate of change of the steering angle ω is 15° / s. These parameters are used to construct a motion model of the target hearing-impaired person to predict their future location.

[0093] When projecting the future trajectory of a target hearing-impaired individual, a time-step approach is used for calculation. Assuming the prediction period is 10 seconds into the future, with a time step of 0.5 seconds, the positions at 20 time points need to be predicted. For each time point t, the current position is calculated based on the previous moment's position, velocity, and direction, combined with motion inertia parameters. Specifically, the position coordinates (xt, yt) at time t are calculated as follows: xt = xt-1 + vx·Δt + 0.5·a·Δt 2 ,yt =yt-1 + vy·Δt + 0.5·a·Δt 2 , where Δt is the time step. Simultaneously update velocity and direction: vx = vx + a·Δt·cos(θ), vy = vy + a·Δt·sin(θ), θ = θ + ω·Δt. In this way, generate a set of first predicted location points P1 = {(x1, y1), (x2, y2), ...,(x20, y20)} for the target hearing-impaired person at multiple future time points.

[0094] For dynamic approach elements, it is necessary to determine their baseline motion trend. This involves obtaining the dominant motion direction φ, motion periodicity characteristics, and motion stability level of the dynamic approach element. For example, a vehicle traveling on a road has a dominant motion direction φ of 60° to the road direction, exhibits moderate periodicity (period T of 5 seconds, amplitude A of 0.5 meters), and a motion stability level of "high" (corresponding to an uncertainty range r of 1 meter).

[0095] The baseline motion trend is determined based on the dominant motion direction, represented as a trajectory with the dominant direction as the reference. For a vehicle, the baseline motion trend is a uniform motion at a speed of v0 = 10 m / s along the direction φ = 60°. However, due to periodicity, the baseline motion trend needs to be corrected. The correction method is to superimpose a periodic offset onto the baseline trajectory. The offset is calculated as a component perpendicular to the dominant direction, with a magnitude of A·sin(2π·t / T). For example, at t = 2.5 seconds, the periodic offset is 0.5·sin(2π·2.5 / 5) = 0.5 meters.

[0096] The uncertainty range *r* is determined based on the motion stability level, representing the possible deviation from the predicted position. For highly stable vehicles, *r* = 1 meter, meaning the actual position is likely within 1 meter of the predicted position. Taking into account the baseline motion trend, periodic corrections, and the uncertainty range, a second predicted position point is generated for each future time *t*.

[0097] (x't, y't): x't = x'0 + v0·t·cos(φ) - A·sin(2π·t / T)·sin(φ), y't = y'0 + v0·t·sin(φ) + A·sin(2π·t / T)·cos(φ). This generates a set of second predicted location points for the dynamic proximity feature at multiple future times.

[0098] P2 = {(x'1, y'1), (x'2, y'2), ..., (x'20, y'20)}.

[0099] Calculating the minimum safety distance requires calculating the spatial distance between corresponding time points in P1 and P2. For time t, the spatial distance dt = √[(xt - x't)]. 2 + (yt - y't) 2 For example, at t=3 seconds, if the predicted location of the target hearing-impaired person is (15.2, 8.7) and the predicted location of the dynamic proximity element is (30.1, 20.3), then the spatial distance between them is 19.8 meters. Iterate through all time points to find the minimum spatial distance. Assuming the minimum distance dmin = 5.2 meters occurs at t=7.5 seconds, this value is taken as the minimum safe distance between the location distributions.

[0100] The above method can accurately predict the future location distribution of the target hearing-impaired person and dynamic access elements, and calculate the minimum safe distance between them, providing a basis for subsequent safety warnings and intervention measures. This method considers the motion inertia of the hearing-impaired person and the motion characteristics of the dynamic access elements, effectively predicting safety risks in various complex scenarios. In practical applications, relevant parameters can be adjusted according to the individual differences of different hearing-impaired persons and the characteristics of different types of dynamic access elements to improve prediction accuracy.

[0101] In one optional implementation, based on the motion pattern characteristics and the risk response level, and combined with a pre-built multi-level signal type screening rule base, non-acoustic perception signal types adapted to the perceptual abilities of the target hearing-impaired person are screened, and the presentation sequence of the non-acoustic perception signals is determined according to the current motion state of the target hearing-impaired person, including:

[0102] A multi-level signal type filtering rule base is established, which includes a first-level filtering rule for risk response level, a second-level filtering rule for motion stability level, and a third-level filtering rule for dominant motion direction.

[0103] The initial candidate signal type set contains all available non-acoustic sensing signal types. The motion stability level and dominant motion direction of the dynamic proximity element are extracted from the motion pattern features. The risk response level is substituted into the first layer of screening rules. Signal types that do not meet the risk response level requirements are excluded from the initial candidate signal type set to obtain the first screening result set.

[0104] Substitute the motion stability level into the second layer of filtering rules, and exclude signal types that do not meet the motion stability level requirements from the first filtering result set to obtain the second filtering result set;

[0105] Substitute the dominant motion direction into the third-level filtering rules, and select the signal type that meets the requirement of the dominant motion direction from the second filtering result set as the finally determined non-acoustic sensing signal type;

[0106] Extract the current movement speed of the target hearing-impaired person from the current movement state, determine the reaction time margin of the target hearing-impaired person based on the current movement speed, calculate the danger signal trigger time based on the reaction time margin and the minimum safe distance, and use the danger signal trigger time as the starting point of the presentation sequence of the non-acoustic sensing signal.

[0107] In the specific implementation process, a multi-level signal type filtering rule base is first established. This rule base contains three layers of filtering rules: a first-level filtering rule based on the risk response level, a second-level filtering rule based on the motion stability level, and a third-level filtering rule based on the dominant motion direction. The first-level filtering rules divide the risk response level into three levels: high, medium, and low, corresponding to different signal type requirements. For example, a high-risk response level requires the signal to have strong attention-awakening characteristics, such as strong vibration or high-brightness flashing light signals; a medium-risk response level can accept moderate-intensity tactile or visual signals; and a low-risk response level can use gentle cue signals, such as slight vibration or soft light cues.

[0108] The second layer of screening rules is based on the level of motion stability, which is divided into three levels: stable, slightly unstable, and severely unstable. For hearing-impaired individuals in a stable state, various types of non-acoustic signals can be selected; for those in a slightly unstable state, strong vibration signals that may interfere with balance should be avoided; and for those in a severely unstable state, visual or gentle tactile signals that do not affect balance should be given priority.

[0109] The third layer of filtering rules filters based on the dominant direction of movement, which includes five basic states: forward, backward, left turn, right turn, and stationary. Different directions correspond to different optimal signal types. For example, the forward state is suitable for visual signals within the field of vision in front, while the left and right turn states are suitable for directional tactile signals on the turning side.

[0110] In practical applications, starting with an initial set of candidate signal types, which includes all available non-acoustic sensory signal types such as vibrational tactile signals, directional tactile signals, high-brightness light signals, color change signals, and icon cue signals, a specific case is considered. Assume a hearing-impaired person is detected walking on a sidewalk at 3.5 km / h, with an electric bicycle approaching from the side and rear at 18 km / h. The calculated risk response level is "high," the motion stability level is "stable," and the dominant motion direction is "forward."

[0111] First, applying the first-level screening rules, based on the "high" risk response level requirement, strong vibration tactile signals, high-brightness flashing light signals, and multimodal combination signals (such as vibration + light signals) are selected from the initial candidate set to form the first screening result set. Next, applying the second-level screening rules, since the hearing-impaired person is in a "stable" movement state, no signal type needs to be eliminated, and the second screening result set is the same as the first screening result set. Finally, applying the third-level screening rules, considering that the dominant movement direction is "forward," high-brightness flashing light signals are preferentially selected from the second screening result set as the final non-acoustic perception signal type, because this signal type can effectively alert the hearing-impaired person without interfering with forward walking.

[0112] After determining the appropriate non-acoustic sensory signal type, the timing of signal presentation needs to be further determined. The movement speed is extracted from the hearing-impaired individual's current motion state; in this case, it is 3.5 km / h. According to ergonomic data, the average reaction time at a walking speed of 3.5 km / h is 1.2 seconds. Considering the potential cognitive characteristics of hearing-impaired individuals, a 30% safety margin is added, resulting in a calculated reaction time margin of 1.56 seconds.

[0113] Meanwhile, based on the relative speed and direction of both parties, the minimum safe distance was calculated to be 3.2 meters. Based on the reaction time margin and the minimum safe distance, the hazard signal trigger point was calculated to be when the distance between the electric bicycle and the hearing-impaired person reaches 5.8 meters. This trigger point serves as the starting point for the presentation sequence of non-acoustic perception signals, ensuring that the hearing-impaired person has sufficient time to react and avoid potential collision risks.

[0114] In practical applications, the intensity and frequency of the signal are dynamically adjusted. For example, as the distance between the electric bicycle and the hearing-impaired person decreases, the flashing frequency of the high-brightness flashing light signal gradually increases from the initial 2 Hz to 4 Hz, and the brightness also increases from the initial 80 candela to 120 candela to enhance the warning effect. This gradual change in signal intensity can more intuitively express the urgency of the risk, helping the hearing-impaired person to more accurately judge potential dangers.

[0115] Through the aforementioned multi-level screening rules and precise timing control, personalized and efficient non-acoustic perception signals can be provided to hearing-impaired individuals, effectively improving their safety in complex traffic environments and reducing the risk of accidents.

[0116] In one optional implementation, dynamically adjusting the directional indication information of the non-acoustic sensing signal based on the real-time positional changes of the dynamic proximity element to guide the target hearing-impaired person to adjust their movement trajectory includes:

[0117] A spatial direction reference coordinate system is established with the current position of the target hearing-impaired person as the origin. The displacement vector between the current position of the dynamic approach element and the previous position of the dynamic approach element is calculated. The approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person is determined based on the displacement vector. The approach azimuth angle is mapped to the spatial direction reference coordinate system to obtain the direction of the hazard source.

[0118] An initial avoidance direction sector is determined based on the direction of the hazard source, wherein the initial avoidance direction sector is a spatial region opposite to the direction of the hazard source.

[0119] Obstacle distribution information of the surrounding environment of the target hearing-impaired person is obtained. The positions of static obstacles within the fan-shaped area of ​​the initial avoidance direction are identified from the obstacle distribution information. Sub-regions in the fan-shaped area of ​​the initial avoidance direction that do not overlap with the positions of the static obstacles are marked as feasible avoidance areas. The center direction vector of the feasible avoidance area is calculated as the preferred avoidance direction.

[0120] Directional indication information is generated based on the preferred avoidance direction. The directional indication information includes the turning angle and turning side relative to the current orientation of the target hearing-impaired person. The directional indication information is encoded into the non-acoustic perception signal. The non-acoustic perception signal containing the directional indication information is output to the target hearing-impaired person through the wearable interactive device to guide the target hearing-impaired person to adjust their movement trajectory.

[0121] In this embodiment, dynamic obstacle avoidance guidance is provided based on the real-time status of the environment surrounding the target hearing-impaired person. Specifically, this is achieved by establishing a spatial reference coordinate system, calculating the movement trajectory of dynamic approach elements, determining the avoidance direction, and generating non-acoustic sensing signals.

[0122] A spatial orientation reference coordinate system is established with the current location of the target hearing-impaired person as the origin. This coordinate system obtains the geographical location information through the positioning module built into the wearable device worn by the target hearing-impaired person, and determines their orientation information by combining the gyroscope and electronic compass. In this coordinate system, the front is defined as 0 degrees, the right side as 90 degrees, the rear as 180 degrees, and the left side as 270 degrees, forming a 360-degree orientation reference system.

[0123] The position change calculation of a dynamic approach element is achieved by comparing its position difference between the current moment and the previous moment. The environmental perception module acquires the position coordinates of the dynamic approach element at time t and time t-1, thereby calculating the displacement vector. The magnitude and direction of the displacement vector reflect the moving speed and direction of the dynamic approach element. Based on the relative relationship between this displacement vector and the position coordinates of the target hearing-impaired person, the approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person is determined.

[0124] The approach azimuth calculation takes into account the relative positional relationship between the target hearing-impaired person's location coordinates and the current position of the dynamic approach element. The azimuth calculated using vectors is mapped to a spatial direction reference coordinate system to obtain the direction of the hazard source. For example, if the calculated approach azimuth is 45 degrees, it means that the hazard source is located 45 degrees to the right front of the target hearing-impaired person.

[0125] When determining the initial avoidance direction sector, the region opposite to the direction of the hazard source is determined in the spatial direction reference coordinate system, using the direction of the hazard source as a reference. If the direction of the hazard source is 45 degrees, then the center direction of the initial avoidance direction sector is 225 degrees. This sector extends 60 degrees to each side of the center direction as an axis, forming a 120-degree sector, that is, the area from 165 degrees to 285 degrees is marked as the initial avoidance direction sector.

[0126] From the obstacle distribution information, the locations of static obstacles within the initial avoidance direction's fan-shaped area are identified. For example, if a wall is identified within the initial avoidance direction's fan-shaped area (165 degrees to 285 degrees) in the range of 200 degrees to 240 degrees, this area is marked as an impassable area. Subsequently, sub-areas within the initial avoidance direction's fan-shaped area that do not overlap with the locations of static obstacles are marked as feasible avoidance areas, namely the 165-degree to 200-degree area and the 240-degree to 285-degree area.

[0127] The center direction vector of the feasible avoidance area is calculated as the preferred avoidance direction. When there are multiple discontinuous feasible avoidance areas, the largest continuous feasible area is selected. In the example above, if the range of the 165° to 200° area is smaller than that of the 240° to 285° area, the latter is selected as the preferred avoidance area, and its center direction is calculated to be approximately 262.5° as the preferred avoidance direction.

[0128] Based on the preferred avoidance direction, directional guidance information is generated. This information includes the turning angle and turning side relative to the target hearing-impaired person's current orientation. If the target hearing-impaired person's current orientation is 0 degrees and the preferred avoidance direction is 262.5 degrees, then a left turn of 82.5 degrees is required. This directional guidance information is encoded as the instruction "Turn left 82 degrees".

[0129] The encoding of non-acoustic sensing signals varies depending on the type of wearable interactive device. For haptic feedback belts, the belt is divided into eight zones, each corresponding to a vibration unit. A command to turn left activates the left-side vibration unit, with the vibration intensity proportional to the turning angle. For visual feedback devices, a left-pointing arrow might be displayed on the smart glasses, with the arrow size reflecting the turning angle.

[0130] Through non-acoustic sensing signals output by wearable interactive devices, the target hearing-impaired person can understand which direction to move in to avoid danger. The system continuously monitors changes in the target hearing-impaired person's position and the movement of dynamic approaching elements, updating directional guidance information in real time. When the target hearing-impaired person successfully avoids danger or the dynamic approaching element no longer poses a threat, obstacle avoidance commands cease, and the system returns to normal monitoring mode.

[0131] In one optional implementation, determining the approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person based on the displacement vector, and mapping the approach azimuth angle to the spatial direction reference coordinate system to obtain the direction of the hazard source includes:

[0132] Obtain the current orientation vector of the target hearing-impaired person, the current orientation vector representing the direction the target hearing-impaired person's body is pointing forward, calculate the vector angle between the displacement vector and the current orientation vector to obtain the relative azimuth angle of the displacement vector relative to the current orientation vector;

[0133] Determine the quadrant interval in which the relative azimuth angle is located, convert the relative azimuth angle into a somatosensory azimuth angle with reference to the body of the target hearing-impaired person based on the quadrant interval, and determine the somatosensory azimuth angle as the approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person;

[0134] An azimuth mapping rule table is established, which stores the mapping relationship between each region identifier of the perceived azimuth and the corresponding direction interval in the spatial direction reference coordinate system. The azimuth mapping rule table is queried according to the region identifier close to the azimuth, and the direction interval of the spatial direction reference coordinate system corresponding to the region identifier is obtained. The center direction angle of the direction interval of the spatial direction reference coordinate system is calculated, and the direction vector corresponding to the center direction angle is determined as the direction of the hazard source.

[0135] This embodiment provides a method for determining the approach azimuth angle of a dynamic approach element relative to a target hearing-impaired person based on a displacement vector, and mapping the approach azimuth angle to a spatial direction reference coordinate system to obtain the direction of the hazard source.

[0136] In practice, the first step is to obtain the current orientation vector of the target hearing-impaired person. This orientation vector represents the direction the person is facing and can be obtained through sensing devices mounted on the person. For example, an electronic compass or gyroscope sensor can be used to detect the orientation. Assuming the obtained current orientation vector is (0.8, 0.6, 0), it indicates that the person is facing a direction approximately 36.9 degrees north on a two-dimensional plane.

[0137] The displacement vector represents the relative position of a dynamic approaching element with respect to the target hearing-impaired person. For example, if an approaching vehicle is detected by environmental perception, its displacement vector relative to the hearing-impaired person is (-0.5, 0.5, 0). This displacement vector indicates that the approaching vehicle is located to the left and in front of the hearing-impaired person.

[0138] After obtaining the displacement vector, the angle between the displacement vector and the current orientation vector is calculated. This is done by dividing the dot product of the two vectors by the product of their respective magnitudes to obtain the cosine of the angle, and then using the inverse cosine function to obtain the actual angle. Using the example data above, the angle between (0.8, 0.6, 0) and (-0.5, 0.5, 0) is approximately 90 degrees. This angle is the relative azimuth angle of the displacement vector with respect to the current orientation vector.

[0139] Next, determine the quadrant of the relative azimuth angle. In practice, with the target hearing-impaired person as the center, their current orientation is directly forward (0 degrees), to the right is 90 degrees, directly behind is 180 degrees, and to the left is 270 degrees, forming a complete 360-degree surround space. The quadrant of the target hearing-impaired person is determined based on the value of the relative azimuth angle. In this example, the relative azimuth angle is 90 degrees, located in the first quadrant, indicating that the dynamic approach element is located to the right of the target hearing-impaired person.

[0140] The relative azimuth is converted into a somatosensory azimuth with the target hearing-impaired person's body as a reference, based on the quadrant interval. This conversion takes into account the natural characteristics of human direction perception, making the output azimuth representation more intuitive for the user. In this example, the relative azimuth is 90 degrees, and the converted somatosensory azimuth is also 90 degrees, corresponding to the "right" direction.

[0141] Once the perceived azimuth is determined, it is used as the approach azimuth of the dynamic approach element relative to the target hearing-impaired person. In this example, the approach azimuth of the dynamic approach element is 90 degrees, indicating that the danger originates from the right side of the hearing-impaired person.

[0142] To map the perceived azimuth to a spatial direction reference coordinate system, an azimuth mapping rule table needs to be established. This table stores the mapping relationship between the regional identifiers of the perceived azimuth and the corresponding directional intervals in the spatial direction reference coordinate system. For example, the perceived azimuth 0-30 degrees and 330-360 degrees are mapped to "front", 30-60 degrees to "right front", 60-120 degrees to "right side", 120-150 degrees to "right rear", 150-210 degrees to "rear", 210-240 degrees to "left rear", 240-300 degrees to "left side", and 300-330 degrees to "left front".

[0143] The azimuth mapping rule table is queried based on the area identifier of the area approaching the azimuth. In this example, the approaching azimuth is 90 degrees, which belongs to the "right" area identifier, corresponding to the 60-120 degree direction interval in the spatial direction reference coordinate system.

[0144] Calculate the central orientation angle of the orientation interval in the spatial orientation reference coordinate system. The central orientation angle of the "right" orientation interval (60-120 degrees) is 90 degrees. Determine the direction vector corresponding to this central orientation angle as the direction of the hazard source. In the standard spatial orientation reference coordinate system, the unit direction vector corresponding to 90 degrees is (0, 1, 0).

[0145] Example of a complete process: Assume the target hearing-impaired person is walking towards north (current orientation vector is (0, 1, 0)). Environmental perception detects a bicycle approaching from the northeast (displacement vector is (0.7, 0.7, 0)). The angle between the displacement vector and the orientation vector is calculated to be approximately 45 degrees, located in the first quadrant. This translates to a perceived azimuth angle of 45 degrees, corresponding to the "right front" area. Consulting the azimuth mapping rule table reveals that "right front" maps to the 30-60 degree range of the spatial direction reference coordinate system, with a central orientation angle of 45 degrees. The corresponding direction vector (0.707, 0.707, 0) is identified as the direction of the hazard source.

[0146] Through the above steps, the displacement information of dynamically approaching elements can be accurately converted into directional information that is easily understood by hearing-impaired individuals, ultimately determining the direction of the hazard source and providing precise directional guidance for subsequent alerts and warnings. This method takes into account the perceptual characteristics of hearing-impaired individuals, enabling more intuitive communication of hazard information and improving the effectiveness of early warnings.

[0147] This invention provides an outdoor sports safety early warning and trajectory tracking protection system for hearing-impaired individuals, comprising:

[0148] The first unit is used to acquire real-time location and motion status information of the target hearing-impaired person in an outdoor environment, continuously monitor the dynamic changes of the environment in which the target hearing-impaired person is located, classify and label the dynamic changes, distinguish between static obstacle elements and dynamic approach elements, and extract the motion pattern features of the dynamic approach elements;

[0149] The second unit is used to generate the future position distribution of the target hearing-impaired person and the dynamic approach element based on the motion state information and the motion pattern characteristics of the dynamic approach element through a trajectory prediction algorithm, calculate the minimum safe distance between the position distributions, and determine the risk of collision and the risk response level when the minimum safe distance is lower than the safety threshold corresponding to the motion pattern characteristics.

[0150] The third unit is used to filter non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person based on the motion pattern characteristics and the risk response level, combined with a pre-built multi-level signal type filtering rule library, and to determine the presentation sequence of the non-acoustic perception signals based on the current motion state of the target hearing-impaired person;

[0151] The fourth unit is used to send sensing commands to the wearable interactive device of the target hearing-impaired person based on the non-acoustic sensing signals, and dynamically adjust the direction indication information of the non-acoustic sensing signals according to the real-time position changes of the dynamic proximity elements, so as to guide the target hearing-impaired person to adjust the movement trajectory.

[0152] A third aspect of the present invention provides an electronic device, comprising:

[0153] processor;

[0154] Memory used to store processor-executable instructions;

[0155] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0156] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0157] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for outdoor sports safety early warning and trajectory tracking protection for hearing-impaired individuals, characterized in that: include: The system acquires real-time location and movement information of the target hearing-impaired individual in an outdoor environment, and continuously monitors changes in dynamic elements of the environment in which the target hearing-impaired individual is located. The dynamic element changes are classified and labeled to distinguish between static obstacle elements and dynamic approach elements, and the motion pattern features of the dynamic approach elements are extracted, including: The dynamic element changes are continuously identified. By analyzing the magnitude and frequency of element position changes within a preset observation period, elements with position changes below a position stability threshold are labeled as static obstacle elements, while elements with position changes exceeding the position stability threshold and whose distance from the target hearing-impaired person shows a decreasing trend are labeled as dynamic approach elements. For objects labeled as dynamic proximity elements, their motion trajectory sequences are established, and pattern clustering analysis is performed on the motion trajectory sequences to identify the dominant motion direction and motion periodicity characteristics of the dynamic proximity elements. Based on the curvature change characteristics and velocity fluctuation characteristics of the motion trajectory sequence, the motion stability level of the dynamic approach element is determined. The motion stability level, the dominant motion direction, and the motion periodicity characteristics together constitute the motion pattern characteristics; Based on the motion state information and the motion pattern characteristics of the dynamic approach elements, a trajectory prediction algorithm is used to generate the future position distribution of the target hearing-impaired person and the dynamic approach elements. The minimum safe distance between the position distributions is calculated. When the minimum safe distance is lower than the safety threshold corresponding to the motion pattern characteristics, a collision risk is determined and a risk response level is identified. Based on the motion pattern characteristics and the risk response level, and combined with a pre-built multi-level signal type screening rule library, non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person are screened, and the presentation sequence of the non-acoustic perception signals is determined according to the current motion state of the target hearing-impaired person. Based on the non-acoustic sensing signal, a sensing command is sent to the wearable interactive device of the target hearing-impaired person, and the directional indication information of the non-acoustic sensing signal is dynamically adjusted according to the real-time position change of the dynamic proximity element to guide the target hearing-impaired person to adjust their movement trajectory.

2. The method according to claim 1, characterized in that, Pattern clustering analysis is performed on the motion trajectory sequence to identify the dominant motion direction and periodicity characteristics of the dynamic approach elements. Based on the curvature change characteristics and velocity fluctuation characteristics of the motion trajectory sequence, the motion stability level of the dynamic approach element is determined as follows: The direction vectors of the trajectory points in the motion trajectory sequence are extracted, the direction vectors are projected onto a preset direction division space, the distribution density of trajectory points in each direction interval is calculated, and the direction interval with the highest distribution density is taken as the dominant motion direction. Calculate the time interval and displacement change between adjacent trajectory points in the motion trajectory sequence to construct a time-displacement correlation sequence. Perform periodic detection on the time-displacement correlation sequence, extract its repetition period, and use the repetition period as a motion periodic feature. The curvature value is calculated for the trajectory segment formed by three consecutive trajectory points in the motion trajectory sequence, a curvature change sequence is generated, the variance and extreme value distribution of the curvature change sequence are statistically analyzed, and the variance and extreme value distribution are normalized and then weighted and summed to determine the trajectory smoothness; The instantaneous velocities corresponding to each trajectory point in the motion trajectory sequence are sampled to generate a velocity change sequence. The standard deviation and coefficient of variation of the velocity change sequence are calculated. After normalizing the standard deviation and the coefficient of variation, a weighted average is performed to determine the velocity stability. When both the trajectory smoothness and the velocity stability exceed the corresponding judgment threshold, the motion stability level is determined to be a high stability level. When the trajectory smoothness or the velocity stability is less than the corresponding determination threshold, the motion stability level is determined to be a low stability level.

3. The method according to claim 2, characterized in that, Based on the motion state information and the motion pattern characteristics of the dynamic proximity element, a trajectory prediction algorithm is used to generate the future position distribution of the target hearing-impaired person and the dynamic proximity element, and the minimum safe distance between the position distributions is calculated, including: Based on the current position coordinates, current velocity vector, and current direction of motion corresponding to the motion state information, and combined with the motion inertial parameters of the target hearing-impaired person, the trajectory evolution process of the target hearing-impaired person in the future predicted time period is deduced, and a set of first predicted position points of the target hearing-impaired person at multiple future moments is generated; The baseline motion trend of the dynamic approach element is determined based on the dominant motion direction. The periodic offset in the baseline motion trend is corrected based on the periodicity characteristics of the motion. The uncertainty range of trajectory prediction is determined based on the motion stability level. Based on the corrected motion trend and the uncertainty range, a second set of predicted position points of the dynamic approach element corresponding to the multiple future moments is generated. Calculate the spatial distance between each predicted location point in the first predicted location point set and the second predicted location point set at the same time. Iterate through all times within the future prediction time period and extract the minimum value of the spatial distance as the minimum safe distance between the location distributions.

4. The method according to claim 1, characterized in that, Based on the motion pattern characteristics and the risk response level, and combined with a pre-built multi-level signal type screening rule base, non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person are screened. Furthermore, based on the current motion state of the target hearing-impaired person, the presentation sequence of the non-acoustic perception signals is determined, including: A multi-level signal type filtering rule base is established, which includes a first-level filtering rule for risk response level, a second-level filtering rule for motion stability level, and a third-level filtering rule for dominant motion direction. The initial candidate signal type set contains all available non-acoustic sensing signal types. The motion stability level and dominant motion direction of the dynamic proximity element are extracted from the motion pattern features. The risk response level is substituted into the first layer of screening rules. Signal types that do not meet the risk response level requirements are excluded from the initial candidate signal type set to obtain the first screening result set. Substitute the motion stability level into the second layer of filtering rules, and exclude signal types that do not meet the motion stability level requirements from the first filtering result set to obtain the second filtering result set; Substitute the dominant motion direction into the third-level filtering rules, and select the signal type that meets the requirement of the dominant motion direction from the second filtering result set as the finally determined non-acoustic sensing signal type; Extract the current movement speed of the target hearing-impaired person from the current movement state, determine the reaction time margin of the target hearing-impaired person based on the current movement speed, calculate the danger signal trigger time based on the reaction time margin and the minimum safe distance, and use the danger signal trigger time as the starting point of the presentation sequence of the non-acoustic sensing signal.

5. The method according to claim 1, characterized in that, Based on the real-time positional changes of the dynamic proximity elements, the directional indication information of the non-acoustic sensing signals is dynamically adjusted to guide the target hearing-impaired person to adjust their movement trajectory, including: A spatial direction reference coordinate system is established with the current position of the target hearing-impaired person as the origin. The displacement vector between the current position of the dynamic approach element and the previous position of the dynamic approach element is calculated. The approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person is determined based on the displacement vector. The approach azimuth angle is mapped to the spatial direction reference coordinate system to obtain the direction of the hazard source. An initial avoidance direction sector is determined based on the direction of the hazard source, wherein the initial avoidance direction sector is a spatial region opposite to the direction of the hazard source. Obstacle distribution information of the surrounding environment of the target hearing-impaired person is obtained. The positions of static obstacles within the fan-shaped area of ​​the initial avoidance direction are identified from the obstacle distribution information. Sub-regions in the fan-shaped area of ​​the initial avoidance direction that do not overlap with the positions of the static obstacles are marked as feasible avoidance areas. The center direction vector of the feasible avoidance area is calculated as the preferred avoidance direction. Directional indication information is generated based on the preferred avoidance direction. The directional indication information includes the turning angle and turning side relative to the current orientation of the target hearing-impaired person. The directional indication information is encoded into the non-acoustic perception signal. The non-acoustic perception signal containing the directional indication information is output to the target hearing-impaired person through the wearable interactive device to guide the target hearing-impaired person to adjust their movement trajectory.

6. The method according to claim 5, characterized in that, The approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person is determined based on the displacement vector. The approach azimuth angle is then mapped to the spatial direction reference coordinate system to obtain the direction of the hazard source, including: Obtain the current orientation vector of the target hearing-impaired person, the current orientation vector representing the direction the target hearing-impaired person's body is pointing forward, calculate the vector angle between the displacement vector and the current orientation vector to obtain the relative azimuth angle of the displacement vector relative to the current orientation vector; Determine the quadrant interval in which the relative azimuth angle is located, convert the relative azimuth angle into a somatosensory azimuth angle with reference to the body of the target hearing-impaired person based on the quadrant interval, and determine the somatosensory azimuth angle as the approach azimuth angle of the dynamic approach element relative to the target hearing-impaired person; An azimuth mapping rule table is established, which stores the mapping relationship between each region identifier of the perceived azimuth and the corresponding direction interval in the spatial direction reference coordinate system. The azimuth mapping rule table is queried according to the region identifier close to the azimuth, and the direction interval of the spatial direction reference coordinate system corresponding to the region identifier is obtained. The center direction angle of the direction interval of the spatial direction reference coordinate system is calculated, and the direction vector corresponding to the center direction angle is determined as the direction of the hazard source.

7. A safety early warning and trajectory tracking protection system for outdoor sports activities of hearing-impaired individuals, used to implement the method as described in any one of claims 1-6, characterized in that, include: The first unit is used to acquire real-time location and motion status information of the target hearing-impaired person in an outdoor environment, continuously monitor the dynamic changes of the environment in which the target hearing-impaired person is located, classify and label the dynamic changes, distinguish between static obstacle elements and dynamic approach elements, and extract the motion pattern features of the dynamic approach elements; The second unit is used to generate the future position distribution of the target hearing-impaired person and the dynamic approach element based on the motion state information and the motion pattern characteristics of the dynamic approach element through a trajectory prediction algorithm, calculate the minimum safe distance between the position distributions, and determine the risk of collision and the risk response level when the minimum safe distance is lower than the safety threshold corresponding to the motion pattern characteristics. The third unit is used to filter non-acoustic perception signal types that are compatible with the perceptual abilities of the target hearing-impaired person based on the motion pattern characteristics and the risk response level, combined with a pre-built multi-level signal type filtering rule library, and to determine the presentation sequence of the non-acoustic perception signals based on the current motion state of the target hearing-impaired person; The fourth unit is used to send sensing commands to the wearable interactive device of the target hearing-impaired person based on the non-acoustic sensing signals, and dynamically adjust the direction indication information of the non-acoustic sensing signals according to the real-time position changes of the dynamic proximity elements, so as to guide the target hearing-impaired person to adjust the movement trajectory.

8. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 6.

Images