Method and system for controlling incoming calls in smartwatches

Abstract

This invention relates to the field of communication terminals and human-computer interaction, and particularly to a method and system for controlling incoming calls for smartwatches. The method for controlling incoming calls for smartwatches includes the following steps: constructing a vertical reference coordinate system, wherein the vertical reference coordinate system is based on a horizontal ground surface, the Z-axis of the vertical reference coordinate system is perpendicular to the horizontal plane and pointing upwards, and the X-axis and Y-axis of the vertical reference coordinate system are both in the horizontal plane and perpendicular to each other. This solution establishes a vertical reference coordinate system based on a horizontal ground surface, sets a specific comparison direction, and uses a gyroscope to collect pose data to determine the watch's orientation; a countdown timer is started when a call comes in, and the call is controlled by monitoring the watch's orientation and accumulating the viewing time. It can adapt to different road conditions and driving habits, accurately identify the intention to answer the call, and solve the problem that smartwatch call answering methods cannot simultaneously address the safety and convenience of driving scenarios without adding new hardware, while also reducing costs and the risk of accidental touches.

Description

Technical Field

[0001] This invention relates to the field of communication terminals and human-computer interaction, and in particular to a method and system for controlling incoming calls for smartwatches. Background Technology

[0002] With the increasing popularity of smartphones, mobile calls have become the core method of daily communication. Smartwatches, with their small size and convenient wearability, are easier to carry around all day compared to phones that need to be held or stored, and are gradually becoming an important auxiliary terminal for smartphones. Utilizing wireless communication technologies such as Bluetooth and NFC, smartwatches can synchronize incoming call signals with the phone in real time. Even if the phone is in a backpack or pocket, where it's difficult to access, users can quickly respond to calls through the smartwatch, effectively avoiding missed calls due to the phone not being nearby, and significantly improving the flexibility of call interaction.

[0003] Traditional smartwatches rely on touchscreen interaction to answer calls. Users must first check the screen to confirm the call, then answer by tapping the answer icon or swiping the touch area. This process is cumbersome, requiring users to pay attention to the screen and use both hands (e.g., one hand holding the watch while the other operates the screen). In situations where hands are busy or distraction is difficult, delays or malfunctions are common, severely impacting user experience. To address this, some smartwatches have simplified the call-answering process by integrating proximity sensors to detect the distance between the watch and the ear. When the watch is detected near the ear—a natural gesture consistent with everyday phone calls—the call is automatically answered, eliminating the need for manual screen touch. This significantly simplifies the process and improves ease of use.

[0004] However, the simplified operation scheme described above still has serious adaptation flaws in driving scenarios. During driving, safe driving is paramount. Users must maintain stable control of the steering wheel with both hands and keep their eyes focused on the road. In this situation, users cannot use their hands for traditional touchscreen operations, nor can they allow their hands to be off the steering wheel for extended periods to perform wrist movements towards the ear. Prolonged hand-off from the steering wheel can easily lead to vehicle instability and potentially cause traffic accidents. At the same time, the safety threshold for looking away from the road while driving is extremely low (typically requiring a single gaze shift of no more than 2 seconds). Users simply cannot keep their eyes on the watch screen for an extended period to confirm the caller's identity. Blindly operating the screen to quickly answer the call will distract driving attention and significantly increase the risk of rear-end collisions and lane departures. Summary of the Invention

[0005] This invention provides a method and system for controlling incoming calls in smartwatches, in order to solve the problem that the incoming call connection method of smartwatches cannot simultaneously take into account the safety and convenience in driving scenarios.

[0006] To solve the above-mentioned technical problems, this application provides the following technical solution:

[0007] A method for controlling incoming calls in a smartwatch includes the following steps:

[0008] A vertical reference coordinate system is constructed, with the horizontal ground as the reference. The Z-axis of the vertical reference coordinate system is perpendicular to the horizontal plane and pointing upwards. The X-axis and Y-axis of the vertical reference coordinate system are both in the horizontal plane and perpendicular to each other. A comparison start point and a comparison end point are selected on the smartwatch, and the direction from the comparison start point to the comparison end point is taken as the comparison direction. A gyroscope is integrated inside the smartwatch, and the gyroscope collects the three-dimensional pose data of the smartwatch in real time. The comparison direction in the three-dimensional pose data is obtained, and the position of the comparison direction in the vertical reference coordinate system is taken as the watch's position. The three-axis acceleration of the smartwatch in the three-dimensional pose data is obtained as the motion acceleration.

[0009] The system acquires the user's location information in real time and processes the location information based on the acquisition time to obtain the user's current movement trajectory. Based on the relationship between the watch's orientation, the rate of change of motion acceleration on the Z-axis, X-axis, and Y-axis, and the rate of change of the slope of the movement trajectory, it determines whether the user is currently in a driving state.

[0010] The smartwatch has preset call reminder duration and call confirmation duration. When the user is currently driving and the smartwatch receives the call information from the receiving terminal, the countdown timer initializes the viewing duration and acquires the watch's location in real time. If the watch's location is within the preset viewing location range before the countdown ends, the acquisition time of the smartwatch's 3D pose data is used as the viewing time, and the viewing duration is accumulated based on the acquisition time of the 3D pose data. If the watch's location is outside the preset viewing location range before the countdown ends, the viewing duration is reset.

[0011] If the countdown timer exceeds the connection confirmation time before it ends, the call will be answered.

[0012] And, a call connection control system for a smartwatch that uses a call connection control method for a smartwatch.

[0013] The basic principle and beneficial effects of this solution are as follows: First, a vertical reference coordinate system is constructed based on the horizontal ground (Z-axis perpendicular to the horizontal plane and upwards, X and Y axes parallel to the horizontal plane). A comparison start point and a comparison end point are selected on the smartwatch. The direction pointing from the comparison start point to the comparison end point is used as the comparison direction. For example, the 6 o'clock position on the watch face is used as the comparison start point, the 12 o'clock position as the comparison end point, and the direction pointing from the 6 o'clock position to the 12 o'clock position is used as the comparison direction. Then, the smartwatch's built-in gyroscope collects three-dimensional pose data in real time, extracts the comparison direction, and maps it to the coordinate system to determine the watch's orientation. Simultaneously, a preset call reminder duration (countdown window) and a call confirmation duration are set. When a call comes in, a countdown begins, and the watch is monitored in real time to ensure it is in the preset viewing position. If within the preset range, the viewing time is accumulated based on the data collection time; if it deviates, the duration is reset. Finally, when the viewing time exceeds the confirmation duration, the call is connected. Users do not need to take their hands off the steering wheel or look away for extended periods while confirming call information and connecting the call, thus resolving the conflict between safety and convenience in driving scenarios. This achieves a balance between the safety requirement of keeping hands on the steering wheel and eyes on the road and the convenience of connecting the call.

[0014] When road conditions are simple and driving is stable, users can easily maintain a stable grip on the steering wheel without consciously adjusting their hand posture, allowing the smartwatch to naturally remain within the preset viewing area. In this case, the viewing time accumulates normally, meeting the user's need for convenient time accumulation to answer calls while driving safely. When road conditions are complex (e.g., traffic congestion requiring frequent minor steering adjustments) or unexpected situations (e.g., avoiding obstacles), users need to change their driving operations to cope with the road conditions. The smartwatch naturally deviates from the preset viewing area with wrist movements, and the viewing time is automatically reset. Without requiring additional user intervention, it naturally transitions to a "no longer waiting for call connection" state, avoiding distractions from continuous waiting or call connection. Furthermore, the user does not need to manually interrupt the judgment process; instead, the natural changes in driving operation are processed into call control operations that better suit the current driving situation. The simple judgment logic can adapt to sudden changes in scenario requirements more quickly, further achieving a balance between safety requirements and convenient call connection. For example, when a user is driving and the vehicle is going through a bumpy road or encountering a sudden situation ahead that requires a quick turn of the steering wheel, the smartwatch will deviate from the preset position according to the wrist movement, and the viewing time will be reset in real time. This will not be misinterpreted as an intention to answer a call, and the user will not need to be distracted by operating the device, thus further balancing driving safety and ease of operation.

[0015] Meanwhile, the dual determination method of preset viewing area range and cumulative viewing time accurately identifies the user's true intention to answer the call. This avoids both accidental triggers caused by momentary touches and missed triggers due to strict judgment when the user has a clear intention. The preset viewing area range requires the user to adjust the smartwatch to a natural viewing posture (excluding unconscious wrist movements). The cumulative viewing time exceeding the connection confirmation time further confirms the user's sustained intention to answer the call (excluding accidental touches from briefly passing by the area). The combination of these two methods forms a double guarantee for intention recognition, which is in line with the user's operating characteristics in driving scenarios. For example, if a user unconsciously rotates their wrist while driving, causing the smartwatch to briefly enter the preset viewing area range, but does not continue until the connection confirmation time, the viewing time will not meet the condition, and the call will not be mistakenly triggered. If the user does have a clear intention to answer the call, they only need to keep the smartwatch in the preset area until the viewing time reaches the target to successfully connect the call without repeating the operation, which is both accurate and convenient.

[0016] Furthermore, the smartwatch's position and three-axis motion acceleration, collected by its built-in gyroscope, serve as the basis for inputting hand gestures to answer calls. This data is also linked to the user's movement trajectory data to determine driving status and filter valid gestures. In driving scenarios, the user's hands are stably gripping the steering wheel, and the watch's position fluctuates within a small, regular range. The rate of change in motion acceleration is strongly correlated with the rate of change in the slope of the movement trajectory. In non-driving scenarios, hand movements are irregular, and data is not linked. Based on this, the user's state can be accurately distinguished. Status recognition and command filtering can be completed without additional hardware, eliminating accidental triggers such as unintentional shaking, thus improving the accuracy of call answering gesture recognition. Simultaneously, it meets the core needs of keeping both hands on the steering wheel while driving and ensuring simple and safe operation, while also considering operational flexibility in non-driving scenarios, adapting to diverse user needs.

[0017] The Z-axis of the vertical reference coordinate system is perpendicular to the horizontal plane, and the X or Y-axis is parallel to the horizontal plane. The orientation of the comparison direction in this coordinate system is only related to the horizontal plane and is independent of the user's natural wrist position angle while driving (such as slightly higher or lower). As long as the watch is positioned within the preset viewing orientation range (for example, when the comparison endpoint is at 12 o'clock and the comparison starting point is at 6 o'clock, the watch face is almost parallel to the Z-axis and 12 o'clock is upward), the viewing time can be accumulated normally. This effect is adapted to the driving habits of different users. For example, some users tend to rest their wrists naturally on the steering wheel while driving. When the watch face is parallel to the horizontal plane, it is difficult for users to observe the road conditions and the watch face at the same time. As the angle between the watch face and the Z-axis decreases, users can observe the road conditions and the watch face more easily. Some users tend to raise their wrists slightly, with the watch face slightly tilted to the sides, while others tend to hold the watch face directly above them. However, as long as both smartwatches meet the preset orientation of being almost parallel to the Z-axis and 12 o'clock upward, the viewing time can be accumulated normally. Users do not need to deliberately change their driving posture to adapt to the device, thus balancing ease of use and driving comfort.

[0018] This solution relies solely on the smartwatch's built-in gyroscope to achieve its functionality, eliminating the need for additional hardware such as proximity sensors. It uses algorithms to perform secondary processing on the 3D pose data collected by the gyroscope (extracting comparison directions and mapping coordinate systems), combined with timing logic to complete control. Compared to solutions requiring additional hardware, this eliminates hardware procurement and circuit design costs, reduces the complexity of equipment production and debugging, improves product cost-effectiveness without increasing user purchase costs, and facilitates large-scale application.

[0019] In summary, this solution establishes a vertical reference coordinate system on a horizontal ground surface, sets a specific comparison direction, and uses a gyroscope to collect pose data to determine the watch's orientation. Upon receiving a call, a countdown timer is initiated, and call connection is controlled by monitoring the watch's orientation and accumulating call duration. This solution adapts to different road conditions and driving habits, accurately recognizes the caller's intent, and solves the problem of smartwatches' call connection methods failing to balance safety and convenience in driving scenarios without requiring additional hardware. It also reduces costs and the risk of accidental touches.

[0020] Furthermore, when the smartwatch starts the call reminder countdown, it acquires the watch's location in real time. Combining the angle between the watch's location and each coordinate axis in the vertical reference coordinate system, it calculates the change angle of the watch's location. If the change angle of the watch's location within the reaction time is greater than the preset adjustment angle, the collection time corresponding to the change in the watch's location angle is integrated into the direction adjustment time period. If the smartwatch starts the call reminder countdown and counts down to the preset adjustment time period, and if the proportion of the direction adjustment time period within the adjustment time period is greater than the preset adjustment proportion after the adjustment time period ends, the call will be answered with a preset message prompt.

[0021] This solution determines call feasibility based on driving intensity. If a user frequently adjusts the steering wheel within the set time, it indicates their hands are unable to maintain stability. Even if the user intends to answer, the call is blocked and a voicemail message is triggered instead, mitigating risk at the root of operational safety. For example, when driving on a series of sharp curves, requiring repeated steering wheel adjustments, allowing a call could easily lead to operational errors due to hand distraction. This solution automatically redirects calls to voicemail when the time spent adjusting direction exceeds the set time, ensuring driving safety while preventing missed calls. Furthermore, it indirectly links driving operations to changes in watch position, reducing user workload and dynamically adapting to different road conditions. Unlike traditional fixed call blocking modes, this solution achieves a reasonable balance between safety and communication needs, and ensures callers receive timely voicemail notifications, reducing interference from repeated calls caused by information asymmetry.

[0022] Furthermore, the smartwatch establishes a communication connection with the driving navigation terminal to obtain driving records; a driving habit model is constructed based on a deep learning model, using the driving speed, steering wheel rotation angle, road driving angle change, and watch position collected at the same time as the same time in the driving records as training data to train the driving habit model;

[0023] After receiving a voicemail message, the smartwatch obtains information about the road, current speed, and changes in driving angle at various points along the road from the navigation terminal. It then inputs the current speed, the road within the preset distance, and the corresponding changes in driving angle at various points along the road into the driving habit model, and outputs the speed at various points along the road, the steering wheel rotation angle, and the watch's position. The smartwatch combines the speed output from the driving habit model with the road information to calculate the driving time period, and then combines the watch's position at various points along the road output from the driving habit model to mark the time period for direction adjustment within the driving time period.

[0024] When the total duration of the direction adjustment period does not exceed the preset adjustment percentage within the driving time, if the similarity between the watch's position change and the preset shaking position change is greater than the preset similarity, the incoming call message will be played; if the similarity is not greater than the preset similarity, the watch's position change will be continuously acquired and compared with the preset shaking position change.

[0025] The driving habit model, combined with data from the navigation terminal, marks the time period for directional adjustments. This not only accurately matches users' driving habits to achieve personalized message playback but also breaks the limitations of relying solely on local smartwatch data. For example, if a user's daily commute route has a series of fixed curves, the driving habit model learns the correlation between steering wheel rotation patterns and watch position changes in that section. When receiving a message while driving on that section again, the model can accurately predict the time period for directional adjustments in the following segment. Provided the user can safely handle the road conditions, the model allows them to listen to the message using specific gestures. This not only improves ease of use but also allows users to access message information without affecting their driving focus. Furthermore, the adaptive nature of the driving habit model allows it to adapt to different users' driving styles and hand gesture habits, expanding the applicability of the solution.

[0026] Furthermore, the smartwatch uses its built-in microphone to collect ambient sound data in real time during the viewing period, calculates the intensity value of the ambient sound data, and combines the collection time to count the percentage of time when the intensity value is higher than the connection noise threshold as the noise percentage before answering the call.

[0027] During the pre-call check period, if the noise level exceeds the preset adjustment level before the call is answered, a preset answer response voice will be sent to the calling terminal after the call is connected, and the preset answer response voice will be played aloud through the built-in speaker; if the noise level does not exceed the preset adjustment level before the call is answered, the call will be connected in a conversational manner.

[0028] The preset playback time of the answer voice message is used as the automatic reply time. During the automatic reply time, if the duration of the ambient sound intensity value being higher than the connection noise threshold exceeds the preset adjustment percentage, or if a call hang-up message is received from the incoming terminal, the call will be transferred to voicemail after the automatic reply time ends.

[0029] If the duration of ambient sound intensity exceeding the connection noise threshold does not exceed the preset adjustment percentage and no call hang-up message is received during the automatic reply period, the real-time collected ambient sound data will be sent to the caller's terminal after the automatic reply period ends.

[0030] This solution switches the call connection mode based on the proportion of ambient sound intensity. It not only optimizes the call experience according to noise levels but also dynamically adapts to the call status. For example, if a user accidentally answers a call while talking to someone nearby, and the ambient noise is high during the time the call is pending, the system will answer with a preset voice prompt and play it aloud, further reminding the user that the call has been connected. Simultaneously, the automatic reply time monitoring prevents communication failures caused by the user not realizing the call has been connected, and allows for timely switching to normal communication once the user notices the call has been connected. This ensures driving safety and avoids unnecessary voicemail redirects.

[0031] Furthermore, the smartwatch stores historical viewing data, including the watch's orientation data and steering wheel rotation angle data collected during the viewing period, combined with the corresponding collection time. The smartwatch also acquires historical viewing data where the steering wheel rotation angle is no greater than a preset adjustment angle, extracts the range of watch orientation changes within each viewing period as a single viewing range, and overlays all single viewing ranges for analysis. It then extracts watch orientations where the percentage of overlaid layers exceeds a preset habitual proportion as habitual orientations, and uses the range formed by this set of habitual orientations as the viewing orientation range. The smartwatch then retrieves historical viewing data according to a preset update cycle and updates the viewing orientation range.

[0032] Iterative updates to the viewing area based on historical data not only improve the accuracy of cumulative viewing time by aligning with user wearing habits, but also enable dynamic self-adaptation of the viewing location. For example, when a user checks an incoming call while driving, they may adjust their viewing location and then to another angle (the user's preferred location). Over time, this solution extracts the user's preferred location from historical data to update the viewing area. This improves the accuracy of recognizing the user's gesture of checking incoming calls, better adapts to the user's answering habits while driving, enhances the user experience, and reduces the need for users to consciously adjust their wrists due to discomfort in the viewing location, further ensuring driving safety. Simultaneously, the preset update cycle design ensures that the viewing area keeps pace with changes in user wearing habits (such as seasonal clothing changes leading to wrist adjustments) while avoiding the resource consumption associated with frequent updates.

[0033] Furthermore, during the countdown to the incoming call reminder, the system obtains information from the navigation terminal regarding the road conditions, current speed, and changes in driving angles at various points along the road within a preset distance. This information is then input into a driving habit model, resulting in the output of predicted driving speed, predicted steering wheel angle, and predicted watch position. The system also obtains the actual steering wheel angle from the navigation terminal. The current speed at the same road location is integrated with the predicted driving speed to calculate the speed interference level within the preset distance. The actual steering wheel angle at the same road location is integrated with the predicted steering wheel angle to calculate the steering interference level within the preset distance. The speed interference level and steering interference level are combined to obtain the driving interference level. The watch position at the same road location is compared with the actual watch position to calculate the user's willingness level. When the viewing time exceeds the connection confirmation time, the system combines the driving interference level with the user's willingness level to connect the call.

[0034] By integrating speed, steering interference, and user intent to determine call connection, this approach not only comprehensively assesses driving status and call-answering needs but also achieves a multi-dimensional balance between safety and demand. For example, when a user encounters a turning section, experiencing significant speed fluctuations and frequent steering wheel movements, resulting in high speed and steering interference, the system detects a substantial change in the user's watch position (i.e., hand-on steering wheel movement) based on driving habits. Combining this prediction with the user's actual watch position accurately identifies the user's strong desire to answer the call despite the complex road conditions. Simultaneously, comparing the predicted normal driving speed and steering wheel movement with the actual situation determines whether the user can maintain a safe driving posture while answering the call—meaning the gesture doesn't compromise driving safety. This method ensures that, for users with good driving skills, road conditions don't limit call-answering needs, while also addressing driving risks, solving the problem of single-dimensional judgments failing to handle complex driving scenarios. Furthermore, it makes the smartwatch's decision-making logic more intelligent and user-friendly, differing from the rigid mode of traditional fixed-condition call connection. Meanwhile, quantifying the degree of interference and the intensity of intention through integral calculation not only improves the accuracy of the judgment but also avoids the uncertainty of subjective judgment.

[0035] Furthermore, when connecting a call, the system combines the degree of interference with the user's willingness to answer. If the degree of interference is lower than a preset interference safety threshold and the user's willingness exceeds a preset connection willingness threshold, the viewing time and call reminder time are extended based on the ratio of the user's willingness to answer to the preset connection willingness threshold.

[0036] Extending the call duration verifies a user's safe driving ability under complex road conditions. For example, if a user is driving on a busy urban-rural mixed road with heavy traffic and initially makes a gesture to answer a call, but there is a sudden risk in the road conditions, extending the viewing time allows the user to maintain a stable watch position within the viewing range and drive without fluctuations. This demonstrates their ability to balance driving and answering calls, preventing safety hazards caused by momentary misjudgments of gestures. Simultaneously, extending the duration further verifies the authenticity of the user's intention to answer. Only if the user maintains the gesture to answer throughout the extended period is a genuine need to answer confirmed, avoiding unnecessary call interruptions. This method avoids blindly blocking important calls while filtering safety risks through time verification. It solves the pain point of traditional fixed-duration methods that cannot distinguish between genuine intention and safety capability versus accidental triggering and risk scenarios, ensuring that call connection is both tailored to user needs and firmly upholds the bottom line of driving safety.

[0037] Furthermore, during the pre-call check period, if the duration of the ambient sound intensity exceeding the connection noise threshold exceeds a preset adjustment percentage, the smartwatch will start a preset response time countdown after the call is connected. If a voice message is received from the calling terminal within the preset response time, the countdown ends and the received voice message is stored locally. If no voice message is received within the response time, a preset dialogue reply message is sent to the calling terminal. If no voice message is received, a preset answer voice message is sent to the calling terminal after the preset response time countdown ends.

[0038] All voice messages received before sending the preset dialogue reply message are integrated into the answer voice message; if a call hang-up message is received before sending the preset dialogue reply message, or if the ambient sound intensity value does not drop below the connection noise threshold, the call will be transferred to voicemail after sending the preset dialogue reply message; if no call hang-up message is received before sending the preset dialogue reply message, and the ambient sound intensity value drops below the connection noise threshold, the answer voice message will be played at a preset acceleration rate, and the playback end time of the answer voice message will be calculated. The sending frequency of the preset dialogue reply message to the incoming call terminal will be adjusted according to the playback end time.

[0039] When a user is talking to a fellow passenger while driving and unintentionally answers a call, this solution first waits for the caller to speak. If the caller remains silent, the call is forwarded to voicemail to avoid wasting call resources. If the caller does speak, a preset reply (or other user-defined voice message) is automatically sent, informing the user of their driving status. Once the environment becomes quiet, the incoming voice message is played at an accelerated pace, allowing the user to quickly grasp the core conversation information while minimizing distraction. Simultaneously, monitoring the user's state before the automatic reply allows for timely forwarding to voicemail if the user is still in an inappropriate conversational situation, preventing communication breakdowns. The accelerated voice playback design solves the problem of users missing conversations due to poor hearing in noisy environments and shortens voice reception time, further ensuring driving safety and making the handling of unintentional calls smarter and safer.

[0040] Furthermore, when calculating the ratio of the user's willingness intensity to the preset connection willingness threshold, if the ratio is greater than the preset upper limit of the willingness ratio, the viewing time and call reminder time will be extended according to the preset upper limit of the willingness ratio.

[0041] After extending the viewing time and call reminder time, the smartwatch obtains the latest driving speed, actual steering wheel rotation angle, watch position and ambient sound data in real time, and recalculates the degree of driving interference.

[0042] If the recalculated level of driving interference is still lower than the preset interference safety threshold, the extended viewing time and call reminder time will be maintained; if the recalculated level of driving interference is higher than the preset interference safety threshold, the viewing time and call reminder time will be adjusted back to the preset viewing time and call reminder time.

[0043] When a user insists on answering calls during complex traffic conditions, the system dynamically re-checks the level of interference while driving and limits the call reminder duration to prevent potential safety risks. For example, if a user receives an urgent call in a congested area and their strong desire to receive it leads to an extended reminder duration, but traffic conditions worsen further during this extended period (e.g., vehicles frequently cutting in), this solution recalculates the level of interference in real time. If the safety risk increases, the original reminder duration is immediately restored, preventing the user from ignoring traffic conditions due to excessive focus on the call. This approach neither absolutely prohibits answering urgent calls nor fails to address sudden changes in traffic conditions, thus resolving the safety hazards of fixed extended reminder durations. Simultaneously, the real-time re-checking method ensures the timeliness of safety assessments, allowing for more flexible duration adjustments. This balances the user's urgent communication needs with driving safety, and makes the smartwatch's call control logic more closely aligned with the complexity of actual driving scenarios, improving the reliability and safety of the solution. Attached Figure Description

[0044] Figure 1 This is a flowchart of a call answering control method for a smartwatch in Example 1;

[0045] Figure 2 This is a flowchart illustrating the formation and updating of the orientation range in Example 1. Detailed Implementation

[0046] The following will describe the concept and technical effects of the present invention clearly and completely with reference to embodiments, so as to fully understand the purpose, features and effects of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention.

[0047] Example 1

[0048] like Figure 1 As shown, the method for controlling incoming calls for a smartwatch includes the following steps:

[0049] A vertical reference coordinate system is constructed, with the horizontal ground as the reference. The Z-axis of the vertical reference coordinate system is perpendicular to the horizontal plane and pointing upwards. The X-axis and Y-axis of the vertical reference coordinate system are both in the horizontal plane and perpendicular to each other. A comparison start point and a comparison end point are selected on the smartwatch. The direction from the comparison start point to the comparison end point is used as the comparison direction. In this embodiment, the 6 o'clock position on the dial surface is used as the comparison start point, the 12 o'clock position as the comparison end point, and the direction from the 6 o'clock position to the 12 o'clock position is used as the comparison direction. A gyroscope (a high-precision gyroscope capable of collecting changes in the watch's pose) is integrated inside the smartwatch. The gyroscope collects the three-dimensional pose data of the smartwatch in real time, and the comparison direction in the three-dimensional pose data is obtained. The position of the comparison direction in the vertical reference coordinate system is used as the watch's orientation. The three-axis acceleration of the smartwatch in the three-dimensional pose data is obtained as the motion acceleration.

[0050] The system acquires the user's location information in real time and processes the location information to obtain the user's current movement trajectory. Based on the relationship between the watch's orientation, the rate of change of motion acceleration on the Z-axis, X-axis, and Y-axis, and the rate of change of the slope of the movement trajectory, it determines whether the user is currently in a driving state. If the user is in a driving state, the system obtains the user's driving time by combining the rate of change of the user's position on the movement trajectory.

[0051] The smartwatch has preset call reminder duration (usually the factory default duration, but can also be set by the user) and connection confirmation duration (usually the factory default duration, but can also be set by the user). When the user is currently driving and the smartwatch receives the call information from the caller, the countdown call reminder duration is initialized to a viewing duration (initial viewing duration is 0), and the watch's location is obtained in real time. If the watch's location is within the preset viewing location range before the countdown ends, the acquisition time of the smartwatch's 3D pose data is used as the viewing time. Combined with the acquisition time of the 3D pose data, the viewing duration is accumulated based on the viewing time. If the viewing duration is greater than the connection confirmation duration, the call is connected; otherwise, the watch's location is obtained in real time before the countdown ends.

[0052] If the watch is not within the preset viewing range before the countdown ends, the viewing time will be reset.

[0053] When the countdown ends, if the duration is no longer than the connection confirmation duration, then turn off the call notification.

[0054] When the user is not currently driving and the smartwatch receives an incoming call notification from the caller, it waits for the user to input a call operation command (entered from the smartwatch screen to answer or turn off the call) before the countdown ends. If the smartwatch does not receive an incoming call notification from the caller, it waits to receive the notification.

[0055] The preset connection duration is determined based on statistical data from simulated safe driving scenarios, including the time it takes for users to check the dashboard to confirm incoming call information, and the sum of the waiting time of 1-3 seconds (administrator setting, assuming the administrator sets it to 2 seconds as the factory setting).

[0056] The system acquires the user's location information in real time (via the positioning device integrated into the smartwatch, or via a vehicle system or smartphone connected to the smartwatch; the positioning device can use a GPS chip). When the user wears the smartwatch, the smartwatch can collect the user's physiological electrical signals. The location information collected while the user is wearing the smartwatch is also the user's location information. The user's location information is combined with the time of acquisition to obtain the user's current movement trajectory. The three-axis acceleration of the smartwatch in the 3D pose data is acquired as the motion acceleration.

[0057] When the user is in a driving posture (such as driving a car), if the user is moving straight (the rate of change of the slope of the trajectory is 0), the user's hand does not move, the watch's position does not change, and the rate of change of motion acceleration along the Z, X, and Y axes is 0; if the user turns, the user's hand turns with the steering wheel, and at this time the rate of change of the watch's position angle, the rate of change of the trajectory slope, and the rate of change of motion acceleration along the Z, X, and Y axes are directly proportional.

[0058] When the user is currently in a riding posture (such as riding a bicycle), if the user is moving straight (the rate of change of the slope of the trajectory is 0), the user's hand does not move, the watch's position does not change, and the rate of change of motion acceleration along the Z, X, and Y axes is 0; if the user turns, the user's hand turns with the steering wheel, at which point the rate of change of the watch's position angle, the rate of change of the trajectory slope, and the rate of change of motion acceleration along the Z and X axes are proportional, while the rate of change along the Y axis is 0.

[0059] By combining the aforementioned driving and riding postures with the movement trajectory and watch position, it is possible to determine whether the user is in a driving state.

[0060] When it is determined that the user is in a driving state, the system combines the speed of position change on the movement trajectory to determine whether the user has stopped moving. If the user has not stopped moving, the time corresponding to each point on the movement trajectory is taken as the driving time. The driving time is combined with the current time to determine whether the user's current state is a driving state (including driving or riding).

[0061] After obtaining the watch's position, it is converted into a vector containing only direction in a vertical reference coordinate system. The administrator simulated a user's driving experience while wearing a watch to determine the observable direction—the position where one can both pay attention to road conditions and observe the watch face through peripheral vision without compromising safety. This observable direction is then converted into a vector containing only direction in a vertical reference coordinate system. All vectors corresponding to observable directions are integrated into an angle range within the vertical reference coordinate system, and this angle range is used as the preset viewing direction range. After obtaining the watch's position, it is converted into a vector containing only direction in a vertical reference coordinate system. If the obtained vector falls within the viewing direction range, the time the watch's position was acquired is used as the viewing time.

[0062] When the smartwatch starts the call reminder countdown, it acquires the watch's position in real time. Combining the angle between the watch's position (i.e., the corresponding vector containing only the position) and each coordinate axis in the vertical reference coordinate system, it calculates the change angle of the watch's position. When the change angle of the watch's position is greater than the preset adjustment angle (determined by statistical data from simulated safe driving scenarios; by simulating the operation of a user turning the steering wheel at a specific angle under different road conditions, collecting the time consumption data corresponding to the operation, and setting the optimal safe time consumption as the factory default value of the reaction time) within the reaction time (the minimum angle of steering wheel rotation when changing the driving direction, determined by statistical data from simulated safe driving scenarios), the collection time corresponding to the change of the watch's position angle is integrated into a direction adjustment time period.

[0063] If the smartwatch starts the call reminder countdown, and the countdown is set to a preset adjustment period, and the proportion of the direction adjustment time within the adjustment period is greater than the preset adjustment proportion after the countdown ends, then the call will be answered with a preset message prompt (after the call is answered, the user's preset message prompt will be sent to the calling terminal, and the received voice information will be retained).

[0064] The preset adjustment duration is determined based on statistical data from simulated user safe driving scenarios: by simulating a scenario where the user does not perform a viewing operation after the incoming call reminder is played, driving hazard probability data is continuously collected under different durations. When the driving hazard probability reaches the minimum threshold preset by the administrator, the corresponding duration is the preset adjustment duration.

[0065] The preset adjustment ratio is determined based on statistical data from simulated safe driving scenarios: by simulating the proportion of time users spend adjusting the steering wheel within a specific time period, the probability data of driving hazards occurring when users perform driving-irrelevant operations (including answering phone calls, making calls, etc.) under different time proportions are continuously collected; when the probability of driving hazards occurring corresponding to the time proportion exceeds the danger probability threshold preset by the administrator, the time proportion is set as the preset adjustment ratio, thereby defining the upper limit of the time proportion during which users can safely perform driving-irrelevant operations.

[0066] The smartwatch establishes a communication connection with the driving navigation terminal to obtain driving records; a driving habit model is built based on a deep learning model, using the driving speed, steering wheel rotation angle, road driving angle change, and watch position collected at the same time as the same time in the driving record as training data to train the driving habit model;

[0067] After receiving a voicemail message, the smartwatch obtains information about the road, current speed, and changes in driving angle at various points along the road from the navigation terminal. It then inputs the current speed, the road within the preset distance, and the corresponding changes in driving angle at various points along the road into the driving habit model, and outputs the speed at various points along the road, the steering wheel rotation angle, and the watch's position. The smartwatch combines the speed output from the driving habit model with the road information to calculate the driving time period, and then combines the watch's position at various points along the road output from the driving habit model to mark the time period for direction adjustment within the driving time period.

[0068] When the total duration of the direction adjustment period does not exceed the preset adjustment percentage within the driving time, if the similarity between the watch's position change and the preset shaking position change is greater than the preset similarity, the incoming call message will be played; if the similarity is not greater than the preset similarity, the watch's position change will be continuously acquired and compared with the preset shaking position change.

[0069] Specifically, after receiving a voicemail message, the smartwatch immediately sends a data acquisition request to the established navigation terminal. The data acquisition request explicitly includes a preset distance parameter (factory default setting, assuming it is set to 2km here). After receiving the request, the navigation terminal retrieves its stored electronic map data and real-time traffic information, filters out the complete road path within the preset distance, and extracts the driving angle change data of each segment of the path (including the angle parameters corresponding to straight sections, curves, slopes, etc.). The road path data within the preset distance and the driving angle change data at each location are packaged and fed back to the smartwatch.

[0070] After receiving feedback data, the smartwatch integrates its own collected current driving speed data and uses the current driving speed, road path data within a preset distance, and driving angle change data at various points on the road as joint input parameters, which are then fed into the trained driving habit model. Based on the user's driving behavior patterns learned during the training phase (combining the correlation logic between driving speed, road angle, steering wheel rotation angle, and watch position), the driving habit model outputs the predicted driving speed, predicted steering wheel rotation angle, and corresponding predicted watch position at various points on the road within the preset distance segment by segment.

[0071] Subsequently, the smartwatch uses the total mileage of the road within a preset distance as a benchmark, combined with the segmented predicted driving speed output by the driving habit model, to calculate the time point for reaching each road segment through the calculation method of "segment mileage ÷ corresponding segment predicted speed," and strings together all time points to form a complete driving time period. At the same time, the smartwatch compares the predicted watch position and corresponding time point output by the driving habit model at various locations on the road with the preset adjustment angle. If the angle of change of the predicted watch position within a certain time period is greater than the preset adjustment angle, then that time period is marked as a direction adjustment time period, ultimately completing the accurate marking of all direction adjustment time periods within the driving time period.

[0072] When a user activates the smartwatch's functions for the first time, the smartwatch will display a prompt to guide the user through the shaking direction input process, while simultaneously showing standardized input steps on the screen (such as "Please complete 2 up-and-down shaking movements within 3 seconds" or "Please complete 2 left-and-right shaking movements within 3 seconds"). The user follows the prompts to perform the target shaking gesture, and the smartwatch, through its built-in gyroscope and accelerometer, synchronously collects complete data on the watch's positional changes during the shaking gesture process (including the positional change trajectory, the positional angle at each moment, and the duration of the change). After continuously collecting 3 sets of valid data, the smartwatch automatically calculates the average of the data and generates "preset shaking positional change baseline data" (including standard trajectory, angle change range, and average shaking speed), which is then stored in the local database.

[0073] Users can initiate the shaking direction adjustment operation through the smartwatch settings interface; the adjustment process is the same as the initial input. The user re-executes the target shaking gesture, the smartwatch collects 3 new sets of data and calculates the average, which overwrites the original baseline data to achieve personalized adjustment (similar to the re-recording logic of the voice assistant wake word, reducing the user's operation threshold).

[0074] When the total duration of the direction adjustment period does not exceed the preset adjustment percentage within the driving time, the smartwatch initiates real-time acquisition of the watch's orientation change (the sampling frequency is set at the factory, assuming it is 10Hz here to ensure data integrity). Using continuous orientation data collected by the gyroscope, the orientation change of adjacent sampling points is calculated. Combined with the sampling time interval, the real-time orientation change speed is obtained by dividing the "orientation change" by the sampling time interval. The real-time change speed is compared with the preset average shaking speed of the orientation change. Only when the real-time change speed is greater than or equal to the average shaking speed will the subsequent similarity matching process begin, filtering out false triggers caused by slow wrist movements.

[0075] Core features (including key node angles of the orientation change trajectory, total angle change range, average shaking speed, and shaking duration) are extracted from the preset shaking orientation change baseline data to construct a feature vector. Simultaneously, features of the same dimension are extracted from the real-time collected watch orientation change data to generate a real-time feature vector. A cosine similarity algorithm is used to calculate the similarity (range 0-1) between the real-time feature vector and the preset feature vector. The core weight allocation is fixed before leaving the factory, assuming the pre-shipment core weight allocation is set as follows: 40% matching degree for key node angles of the trajectory, 30% matching degree for the total angle change range, 20% deviation for average shaking speed, and 10% deviation for shaking duration.

[0076] When comparing the real-time collected watch position change data with the preset shaking position change data, if the calculated similarity is greater than the preset similarity threshold (the default setting before leaving the factory, assuming the default setting is 0.7, and users can adjust it in the settings interface), a message playback command is generated to control the smartwatch to play the incoming call message through the speaker; if the similarity is less than or equal to the preset similarity threshold, the watch position change data is continuously collected. For each complete set of data collected (the duration is consistent with the preset shaking duration), the above preprocessing and similarity calculation steps are repeated until a shaking gesture that meets the conditions is collected, or the total duration of the non-direction adjustment time period does not meet the requirement that the proportion of the total duration of the non-direction adjustment time period in the driving time period is not greater than the preset adjustment proportion.

[0077] The smartwatch uses its built-in microphone to collect ambient sound data in real time during the viewing period, calculates the intensity value of the ambient sound data, and combines the collection time to count the percentage of time when the intensity value is higher than the connection noise threshold as the noise percentage before answering the call.

[0078] During the pre-call screening period, if the noise level exceeds the preset adjustment level, a preset answer voice message (set by the user according to their personal needs before use) will be sent to the calling terminal after the call is connected, and the preset answer voice message will be played aloud through the built-in speaker. If the noise level does not exceed the preset adjustment level, the call will be connected in a conversational manner (meaning that after the call is connected, the smartwatch and the calling terminal establish a two-way real-time audio transmission channel to realize instant voice interaction between the user and the caller).

[0079] The preset playback time of the answer voice message is used as the automatic reply time. During the automatic reply time, if the duration of the ambient sound intensity value being higher than the connection noise threshold exceeds the preset adjustment percentage, or if a call hang-up message is received from the incoming terminal, the call will be transferred to voicemail after the automatic reply time ends.

[0080] If the duration of ambient sound intensity exceeding the connection noise threshold does not exceed the preset adjustment percentage and no call hang-up message is received during the automatic reply period, the real-time collected ambient sound data will be sent to the caller's terminal after the automatic reply period ends.

[0081] like Figure 2 As shown, the smartwatch stores historical viewing data, including watch orientation data and steering wheel rotation angle data collected during the viewing period, combined with the corresponding collection time. The smartwatch acquires historical viewing data where the steering wheel rotation angle is no greater than a preset adjustment angle. It extracts the range of watch orientation changes within each viewing period as a single viewing range. All single viewing ranges are overlaid and analyzed. The watch orientation where the percentage of overlaid layers exceeds a preset habitual proportion (a default value set at the factory, generally above 70%, assumed to be 75% here) is selected as the habitual orientation. The range formed by this set of habitual orientations is then used as the viewing orientation range. The smartwatch updates the viewing orientation range based on the historical viewing data according to a preset update cycle (a default value set at the factory, which users can modify after purchase).

[0082] During the countdown to the incoming call reminder, the system obtains information from the navigation terminal regarding the road conditions within a preset distance, the current speed, and changes in driving angles at various points along the road. It also inputs a driving habit model and outputs predicted driving speed, predicted steering wheel rotation angle, and predicted watch position. The system also obtains the actual steering wheel rotation angle from the navigation terminal.

[0083] The speed interference level within a preset distance is obtained by integrating the current speed and the predicted driving speed at the same road location; the steering interference level within a preset distance is obtained by integrating the actual steering wheel rotation angle and the predicted steering wheel rotation angle at the same road location; the driving interference level is obtained by combining the speed interference level and the steering interference level; and the user's intention intensity is obtained by comparing and integrating the watch position at the same road location with the actual watch position.

[0084] When the viewing time exceeds the call confirmation time, the call will be answered based on the degree of driving interference and the user's willingness. Specifically, when answering a call based on the degree of driving interference and the user's willingness, if the degree of driving interference is lower than a preset interference safety threshold and the user's willingness exceeds a preset call connection willingness threshold, the viewing time and call reminder time will be extended according to the ratio of the user's willingness to the preset call connection willingness threshold.

[0085] More specifically, the total steering angle within a preset distance is calculated using a block-level formula. , ,in, This represents the change in driving angle corresponding to the i-th road position within a preset distance. It represents cumulative calculation; it quantifies the overall steering demand within a preset distance by superimposing changes in driving angles at various points along the road.

[0086] The smartwatch has a pre-stored driving difficulty map, which presets the driving speed weights for different total steering angle ranges. Weight of steering wheel rotation angle And satisfy ; calculate the Match the steering angle range with the driving difficulty map table and extract the corresponding values. and This enables differentiated allocation of interference factor weights based on driving difficulty.

[0087] The administrator tracks driving accidents under different road conditions and total steering angles, categorizing them into speed-related and turning-related causes based on their occurrence. The administrator then calculates the percentage of speed-related and turning-related accidents under each total steering angle and sets a weighting for driving speed based on this percentage. Weight of steering wheel rotation angle Weighting the same driving speed Weight of steering wheel rotation angle The corresponding total steering angles are combined into a total steering angle range, and different total steering angle ranges are weighted according to their corresponding driving speeds. Weight of steering wheel rotation angle Integrated into a driving difficulty mapping table.

[0088] When obtaining the degree of speed interference, the integral result of the speed difference is calculated by the block-level formula, as shown in formula (1) below:

[0089] (1),

[0090] This represents the integral result of the velocity difference;

[0091] S represents the preset distance;

[0092] This indicates the current speed at road location s;

[0093] This represents the predicted driving speed at road location s;

[0094] This indicates the integral operation along a preset distance.

[0095] The smartwatch contains a mapping table of speed difference integral ranges and speed interference values, determined by the administrator based on statistical data from simulated safe driving scenarios. The speed disturbance value is obtained by matching the integral range of the speed difference. Quantify the degree of deviation between the actual speed and the predicted speed (i.e., the degree of speed interference), speed interference value for and The difference and The ratio of .

[0096] The integral result of the steering angle difference is calculated using the block-level formula, as shown in the following formula (2):

[0097] (2),

[0098] This represents the integral result of the steering angle difference;

[0099] This indicates the actual steering wheel rotation angle at road location s;

[0100] This represents the predicted steering wheel angle at road location s;

[0101] The smartwatch has a pre-set mapping table between the steering wheel rotation angle range and the steering interference value. This table is determined by an administrator based on statistical data from simulated safe driving scenarios, and includes the integral range of the steering angle difference and the steering interference value. The mapping table; The steering interference value is obtained by matching the steering wheel rotation angle range. This quantifies the degree of deviation between the actual steering and the predicted steering (i.e., the degree of steering interference). Steering interference value. for and The difference and The ratio of .

[0102] The driving disturbance value is calculated by combining the block-level formula with the weight allocation, as shown in the following formula (3):

[0103] (3),

[0104] D represents the driving disturbance value, which combines speed disturbance and steering disturbance with the weights corresponding to driving difficulty to accurately quantify the overall disturbance level during the driving process.

[0105] The intensity of user willingness is calculated using a block-level formula, as shown in formula (4) below:

[0106] (4),

[0107] I represents the strength of the user's willingness;

[0108] This indicates the predicted watch orientation at road location s;

[0109] Indicates the actual position of the watch at road location 's';

[0110] The larger the score (I), the more significant the deviation between the actual watch position and the predicted position, and the stronger the user's willingness to answer the call.

[0111] When D is lower than the interference safety threshold determined by the administrator based on statistical data from simulated user safe driving scenarios. And I exceeds the preset threshold for willingness to connect. At that time, the duration extension ratio is calculated using a block-level formula: Where k represents the duration extension ratio; the viewing time is extended to (original viewing time × k), and the call reminder time is (call reminder time + original viewing time × k).

[0112] And, a call connection control system for a smartwatch that uses a call connection control method for a smartwatch.

[0113] In practice

[0114] The specific implementation process of the method by which driver A uses a smartwatch to answer calls and control the system.

[0115] Driver A uses a smartwatch that supports this invention for commuting. The watch has established communication with the vehicle's navigation terminal. Its driving habit model has been trained using A's driving records for the past 3 months (including speed, steering wheel angle, road angle, and watch position data). Parameters such as the 30-second call reminder duration and the 4-second call confirmation duration are all personalized according to A's habits.

[0116] While driving, the watch establishes a vertical reference coordinate system, using the direction from 6 o'clock to 12 o'clock on the dial as the comparison direction. It uses a built-in high-precision gyroscope to collect three-dimensional pose data in real time, converting it into a vector containing only orientation as the watch's position. Simultaneously, the watch continuously stores historical viewing data and updates the preset viewing orientation range weekly to ensure it adapts to driver A's preferred wrist position.

[0117] When A drives onto a main urban road, an incoming call is transmitted to the watch, which then starts a 30-second call reminder countdown, initially displaying a viewing duration of 0. At this time, A's wrist is placed naturally, and the watch's position is within a preset range, with the viewing duration continuously accumulating; if A turns and causes the position to deviate, the display is reset.

[0118] Simultaneously, the watch collects ambient sound through the microphone and calculates the noise level before the call is answered. Assuming that the navigation terminal obtains the road path 2km ahead, the current vehicle speed of 60km / h, changes in road angle, and actual steering wheel angle data, the watch inputs the driving habit model to output prediction parameters, calculates the driving time period, and marks the direction adjustment time period corresponding to the curve.

[0119] The watch accumulates road angle changes to obtain the total steering angle, matches it with a driving difficulty mapping table to obtain weights, calculates the cumulative difference between speed and steering, matches it with a mapping table to obtain interference values, and calculates the driving interference value by weighted summation; at the same time, the watch's orientation deviation is calculated to obtain the user's intention intensity.

[0120] Assuming that, based on statistics, noise accounts for 45% of the time before answering the call and direction adjustment time accounts for 30%, neither exceeds the preset adjustment rate of 60%. If the viewing time reaches 5 seconds (exceeding the connection confirmation time by 4 seconds), and the driving interference value is below the threshold and the willingness intensity meets the standard, the watch extends the viewing time to 7 seconds, the reminder time to 33 seconds, and rechecks to confirm safety.

[0121] After observing the incoming call notification, A makes two pre-defined up-and-down shaking gestures. The watch samples at 10Hz, extracts the gesture features, and calculates the similarity with pre-defined baseline data. The result is 0.82, exceeding the threshold of 0.7, confirming the gesture's validity. At this point, the ambient sound level drops below the noise threshold, and the watch plays the answer message at 1.2x speed, adjusting the timing of the response message.

[0122] When the extended viewing time reaches 7 seconds, the watch generates a call connection command, establishing a two-way real-time audio channel to enable secure communication between A and the caller. If sudden changes in road conditions cause interference levels to exceed the limit, the watch immediately restores the original time setting to ensure driving safety, without affecting A's driving concentration throughout the process.

[0123] Example 2

[0124] The only difference between this embodiment and embodiment 1 is that, during the viewing period before the call is connected, if the duration of the ambient sound intensity value being higher than the connection noise threshold exceeds the preset adjustment percentage, the smartwatch will start a countdown for the response time after the call is connected.

[0125] If a voice message is received from the incoming call terminal within the response time, the countdown will stop and the received voice message will be stored locally. If no voice message is received, after the countdown ends, a preset answer voice message will be continuously sent to the incoming call terminal at a preset fixed interval (the preset answer voice message is a series of ordered sequences, and the duration of each message and the total number of messages in the sequence are pre-stored locally).

[0126] All voice messages received within the preset response time are integrated into the received voice messages according to the receiving time sequence. The original playback duration of the received voice messages is calculated. Combined with the preset acceleration rate (stored locally and adjustable by the user), the actual playback end time of the received voice messages is calculated using the formula "Playback end time = Current time + (Original playback duration ÷ Acceleration rate)".

[0127] If a call is disconnected before the preset dialogue reply sequence is sent, or if the ambient sound intensity value does not drop below the connection noise threshold, the call will be switched to voicemail after the preset dialogue reply sequence is sent.

[0128] If no call hang-up message is received before sending the preset dialogue reply information sequence and the ambient sound intensity value drops below the connection noise threshold, then the number of messages that have not been sent in the preset dialogue reply information sequence is extracted. If no voice message is received from the incoming terminal within the reaction time (i.e., the caller does not explain within the reaction time, such as when they stop speaking), the preset dialogue reply information sequence is sent to the incoming terminal one by one.

[0129] The start time of sending the preset dialogue reply information sequence is taken as the end time of receiving the voice message, and the time when the voice message is first received is taken as the start time. The duration of the voice message is calculated based on the end time and the start time.

[0130] The playback start time for the received voice message is defined as the time when the ambient sound intensity drops below the connection noise threshold. The voice message is played at a preset acceleration rate at this start time. The playback end time is calculated as "playback start time + duration of the received voice message ÷ preset acceleration rate". The reply start time is defined as the time when the sequence of dialogue reply messages is sent to the incoming call terminal. The sequence of dialogue reply messages is sent to the incoming call terminal at the factory-set default sending frequency. The predicted reply end time is calculated as "sending frequency × total number of messages in the sequence + reply start time".

[0131] If the playback end time is less than or equal to the predicted response end time, the sending frequency will not be changed.

[0132] If the playback end time is greater than the predicted reply end time, the sending interval time is calculated as "(playback end time - reply start time) ÷ total number of messages in the sequence". If the interval time is not less than the preset minimum interval (the default value is set before leaving the factory, based on experimental testing, and determined according to the fastest speech speed that does not affect the caller's ability to hear the dialogue reply information), the dialogue reply information sequence is sent to the caller's terminal at the sending interval time; if the interval time is less than the preset minimum interval, the dialogue reply information sequence is sent to the caller's terminal at the minimum interval.

[0133] When calculating the ratio of user willingness intensity to a preset connection willingness threshold, if the ratio is greater than the preset upper limit of the willingness ratio, the viewing time and call reminder time will be extended according to the preset upper limit of the willingness ratio. After extending the viewing time and call reminder time, the smartwatch obtains the latest driving speed, actual steering wheel rotation angle, watch position, and ambient sound data in real time to recalculate the degree of driving interference. If the recalculated degree of driving interference is still lower than the preset interference safety threshold, the extended viewing time and call reminder time will be maintained. If the recalculated degree of driving interference is higher than the preset interference safety threshold, the viewing time and call reminder time will be adjusted back to the preset viewing time and call reminder time.

[0134] In practice

[0135] Driver A wears a smartwatch that supports this invention while driving. The watch communicates stably with the vehicle's navigation terminal. The driving habit model has been trained with long-term data. Parameters such as call reminder duration and connection confirmation duration are set according to his / her habits. The preset dialogue reply voice sequence, minimum sending interval, and willingness ratio upper limit have been configured.

[0136] While driving, driver A's watch receives an incoming call notification. The watch starts a call reminder countdown, initializes the call duration, and simultaneously collects ambient sound data. Statistics show that the noise level before the call is answered exceeds a preset threshold, triggering the core control logic.

[0137] After an incoming call is connected, the watch starts a preset response time countdown. If no voice is received from the caller, the watch sends a preset answer response sequence at fixed intervals after the countdown ends. The watch integrates the scattered voices received within the response time into the answer voice, calculates the original playback duration and combines it with a preset acceleration rate. After the ambient noise drops to a threshold, the watch calculates the actual playback end time.

[0138] The watch uses the start time of the reply sequence as the reply start time and calculates the predicted reply end time. If the actual playback end time is later, a new sending interval is calculated according to the formula, ensuring that the remaining replies are sent after the minimum interval is not less than the initial interval.

[0139] The system simultaneously calculates the ratio of user willingness intensity to a threshold. If the ratio exceeds the upper limit, the viewing time and call reminder time are extended by the upper limit, and the degree of driving interference is rechecked. If the threshold is met, the extended time is maintained. During this period, driver A makes a preset shaking gesture, and if the similarity meets the standard, the willingness to answer the call is confirmed to be valid.

[0140] If the extended viewing time exceeds the confirmed call duration, and the ambient noise level is within acceptable limits with no missed or disconnected messages, the watch establishes a two-way real-time audio channel. If unexpected road conditions cause excessive interference, the original duration is immediately restored; if the call connection conditions are not met, the watch sends a reply voice message and then switches to voicemail, balancing safety and communication needs.

[0141] The above are merely embodiments of the present invention. The invention is not limited to the fields covered by these embodiments. Commonly known structures and characteristics in the solutions are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are able to access all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A method for controlling incoming calls in a smartwatch, characterized in that, Includes the following steps: A vertical reference coordinate system is constructed, with the horizontal ground as the reference. The Z-axis of the vertical reference coordinate system is perpendicular to the horizontal plane and pointing upwards. The X-axis and Y-axis of the vertical reference coordinate system are both in the horizontal plane and perpendicular to each other. A comparison start point and a comparison end point are selected on the smartwatch, and the direction from the comparison start point to the comparison end point is taken as the comparison direction. A gyroscope is integrated inside the smartwatch, and the gyroscope collects the three-dimensional pose data of the smartwatch in real time. The comparison direction in the three-dimensional pose data is obtained, and the position of the comparison direction in the vertical reference coordinate system is taken as the watch's position. The three-axis acceleration of the smartwatch in the 3D pose data is obtained as the motion acceleration; The system acquires the user's location information in real time and processes the location information based on the acquisition time to obtain the user's current movement trajectory. When the slope of the movement trajectory changes at a rate of 0, the user's hand does not move, the watch's orientation does not change, and the acceleration changes at a rate of 0 on the Z, X, and Y axes, the system determines that the user is in a straight-line driving state. When the user's hand turns with the steering wheel, the speed at which the watch's orientation changes, the speed at which the trajectory slope changes, and the speed at which the acceleration changes in the Z, X, and Y axes are all proportional to the speed at which the user is in a turning driving state. The smartwatch has preset call notification duration and connection confirmation duration; When the user is currently driving and the smartwatch receives incoming call information from the caller, it counts down the call reminder duration, initializes the viewing duration, and obtains the watch's location in real time. If the watch is within the preset viewing range before the countdown ends, the viewing time will be the acquisition time of the smartwatch's 3D pose data. The viewing duration will be accumulated based on the acquisition time of the 3D pose data. If the watch is outside the preset viewing range before the countdown ends, the viewing duration will be reset. If the countdown timer exceeds the connection confirmation time before it ends, the call will be answered.

2. The call connection control method for a smartwatch according to claim 1, characterized in that: When the smartwatch starts the countdown for incoming call reminders, the watch's position is acquired in real time. The angle between the watch's position and each coordinate axis in the vertical reference coordinate system is used to calculate the change angle of the watch's position. When the change angle of the watch's position within the preset reaction time is greater than the preset adjustment angle, the acquisition time corresponding to the change of the watch's position angle is integrated into the direction adjustment time period. If the smartwatch starts the call reminder countdown, and the countdown is set to a preset adjustment period, and the proportion of the direction adjustment time within the adjustment period after the countdown ends is greater than the preset adjustment proportion, then the call will be answered with a preset message prompt.

3. The call connection control method for a smartwatch according to claim 2, characterized in that: The smartwatch establishes a communication connection with the navigation terminal to obtain driving records; A driving habit model is built based on a deep learning model. The driving speed, steering wheel rotation angle, road driving angle change, and watch position collected by the smartwatch at the same time as the same collection time in the driving record are used as training data to train the driving habit model. After receiving a voicemail message, the smartwatch obtains the road within a preset distance, current speed, and changes in driving angle at various points on the road from the driving navigation terminal. It then inputs the current speed, the road within the preset distance, and the corresponding changes in driving angle at various points on the road into the driving habit model, and outputs the speed at various points on the road, the steering wheel rotation angle, and the watch's position. The driving habit model outputs speed and road conditions to calculate driving time, and then the driving habit model outputs watch positions at various points on the road to mark the time period for direction adjustment. When the total duration of the direction adjustment period is no greater than the preset adjustment percentage within the driving period, if the similarity between the watch's position change and the preset shaking position change is greater than the preset similarity, then the incoming call message will be played. If the similarity is not greater than the preset similarity, the watch's positional changes will be continuously acquired and compared with the preset shaking positional changes.

4. The call connection control method for a smartwatch according to claim 3, characterized in that: The smartwatch uses its built-in microphone to collect ambient sound data in real time during the viewing period, calculates the intensity value of the ambient sound data, and combines the collection time to count the percentage of time when the intensity value is higher than the connection noise threshold as the noise percentage before answering the call. During the pre-call check period, if the noise level exceeds the preset adjustment level before the call is answered, a preset answer response voice will be sent to the calling terminal after the call is connected, and the preset answer response voice will be played aloud through the built-in speaker; if the noise level does not exceed the preset adjustment level before the call is answered, the call will be connected in a conversational manner. The preset playback time of the answer voice message is used as the automatic reply time. During the automatic reply time, if the duration of the ambient sound intensity value being higher than the connection noise threshold exceeds the preset adjustment percentage, or if a call hang-up message is received from the incoming terminal, the call will be transferred to voicemail after the automatic reply time ends. If the duration of ambient sound intensity exceeding the connection noise threshold does not exceed the preset adjustment percentage and no call hang-up message is received during the automatic reply period, the real-time collected ambient sound data will be sent to the caller's terminal after the automatic reply period ends.

5. The call connection control method for a smartwatch according to claim 3, characterized in that: The smartwatch will combine the watch's position data and steering wheel rotation angle data collected during the viewing period with the corresponding collection time to store historical viewing data; The smartwatch acquires historical viewing data where the steering wheel rotation angle is no greater than the preset adjustment angle. It extracts the range of change in the watch's orientation within each viewing period as the single viewing range. All single viewing ranges are overlaid and analyzed. The watch orientations where the number of overlay layers exceeds the preset habitual proportion in the total number of overlays are extracted as habitual orientations. The orientation range formed by the set of habitual orientations is used as the viewing orientation range. The smartwatch retrieves historical viewing data according to a preset update cycle and updates the viewing location range accordingly.

6. The call connection control method for a smartwatch according to claim 5, characterized in that: During the countdown of incoming call reminders, the system obtains information from the navigation terminal about the road, current speed, and changes in driving angle at various points along the road within a preset distance. It also inputs the driving habit model and outputs the predicted driving speed, the predicted steering wheel rotation angle, and the predicted watch position. Obtain the actual steering wheel rotation angle from the driving navigation terminal; The speed interference level within a preset distance is obtained by integrating the current speed and the predicted driving speed at the same road location; the steering interference level within a preset distance is obtained by integrating the actual steering wheel rotation angle and the predicted steering wheel rotation angle at the same road location. The degree of driving interference is obtained by combining the degree of speed interference and the degree of steering interference; the intensity of user intention is obtained by comparing the watch position corresponding to the same road location with the actual watch position and calculating the integral. When the viewing duration exceeds the connection confirmation duration, the call will be connected based on the degree of driving interference and the user's willingness.

7. The call connection control method for a smartwatch according to claim 6, characterized in that: When connecting a call, the system combines the degree of interference with the user's willingness to connect. If the degree of interference is lower than a preset interference safety threshold and the user's willingness exceeds a preset connection willingness threshold, the viewing time and call reminder time will be extended based on the ratio of the user's willingness to connect to the preset connection willingness threshold.

8. The call connection control method for a smartwatch according to claim 4, characterized in that: During the pre-call check period, if the duration of the ambient sound intensity value being higher than the connection noise threshold exceeds the preset adjustment percentage, the smartwatch will start a preset response time countdown after the call is connected. If a voice message is received from the calling terminal within the preset response time, the countdown will end and the received voice message will be stored locally. If no voice message is received within the response time, a preset dialogue reply message will be sent to the calling terminal. If no voice message is received, a preset answer voice message will be sent to the caller's terminal after the countdown of the preset response time has ended; All voice messages received before sending the preset dialogue reply message are integrated into the answer voice message; If a call is disconnected before the preset dialogue reply message is sent, or if the ambient sound intensity value does not drop below the connection noise threshold, the call will be transferred to voicemail after the preset dialogue reply message is sent. If no call hang-up message is received before sending the preset dialogue reply message, and the ambient sound intensity value drops below the connection noise threshold, the answer voice message is played at a preset acceleration rate. At the same time, the playback end time of the answer voice message is calculated, and the sending frequency of the preset dialogue reply message to the caller terminal is adjusted according to the playback end time.

9. The call connection control method for a smartwatch according to claim 7, characterized in that: When calculating the ratio of user willingness intensity to preset connection willingness threshold, if the ratio is greater than the preset upper limit of willingness ratio, the viewing time and call reminder time will be extended according to the preset upper limit of willingness ratio. After extending the viewing time and call reminder time, the smartwatch obtains the latest driving speed, actual steering wheel rotation angle, watch position and ambient sound data in real time, and recalculates the degree of driving interference. If the recalculated level of driving interference is still lower than the preset interference safety threshold, the extended viewing time and call reminder time will be maintained. If the recalculated level of driving interference exceeds the preset interference safety threshold, the viewing duration and call reminder duration will be adjusted back to the preset viewing duration and call reminder duration.

10. A call connection control system for a smartwatch, characterized in that, The method for controlling incoming calls for a smartwatch according to any one of claims 1-9 was used.

Images