Calorie detection method based on wearable device, wearable device and medium

Abstract

The invention discloses a calorie detection method based on a wearable device, the wearable device and a medium, relates to the technical field of wearable devices, and discloses the calorie detection method based on the wearable device, and the method comprises the following steps: according to inertial data of an inertial sensor and sensor data of other sensors except the inertial sensor, determining a calorie value of the inertial sensor; judging whether a calorie identification process is triggered or not; if yes, the current calorie is determined according to the food type and the food component in the food image collected by the image collection module. Non-sensitive triggering of the calorie recognition process is realized by combining different dimension parameters, complexity of manual operation of a user is avoided, the accuracy of scene recognition is improved, false triggering of a non-eating scene is reduced, calorie detection can be completed without complex operation in daily diet of the user in a portable use scene of the wearable device, and the user experience is improved. And the use convenience and the data accuracy are improved.

Description

Technical Field

[0001] This application relates to the field of wearable device technology, and in particular to a calorie detection method, wearable device, and medium based on wearable devices. Background Technology

[0002] When identifying dietary calorie intake using wearable devices, the process requires clicking the calorie recognition function on the device itself, or accessing the calorie recognition function through a connected terminal such as a mobile phone, and then identifying the calories in the food based on the captured images. Therefore, current dietary calorie recognition methods suffer from a cumbersome process.

[0003] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a calorie detection method, wearable device and medium based on wearable devices, which aims to solve the technical problem of calorie identification requiring manual operation, which is cumbersome and has low identification efficiency.

[0005] To achieve the above objectives, this application proposes a calorie detection method based on a wearable device, the method comprising: Based on the inertial data from the inertial sensor and the sensor data from other sensors besides the inertial sensor, determine whether to trigger the calorie recognition process; If so, the current calorie count is determined based on the type and portion size of the food in the food image captured by the image acquisition module.

[0006] In one embodiment, the other sensors include at least a distance sensor and an ambient light sensor, and the step of determining whether to trigger the calorie recognition process based on inertial data from the inertial sensor and sensor data from the other sensors besides the inertial sensor includes: If the head posture corresponding to the inertial data meets the preset posture, the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor are obtained. Based on the multimodal fusion results of the inertial data, the distance parameters, and the ambient light data, it is determined whether to trigger the calorie recognition process.

[0007] In one embodiment, before the steps of acquiring the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor, the calorie detection method based on the wearable device further includes: Acquire target inertial data collected by a wristband device connected to a wearable device, and determine the hand posture corresponding to the target inertial data; If the hand posture meets the preset posture, the steps of obtaining the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor are executed; After the steps of acquiring the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor if the head posture corresponding to the inertial data satisfies the preset posture, the calorie detection method based on wearable devices further includes: Based on the multimodal fusion results of the inertial data, the target inertial data, the distance parameters, and the ambient light data, it is determined whether to trigger the calorie recognition process.

[0008] In one embodiment, after the step of determining the current calories based on the food type and portion size in the food image acquired by the image acquisition module, the calorie detection method based on the wearable device further includes: Based on the current calories and the historical calories consumed within the preset time period, the total calories consumed are determined; Based on the total calories ingested and the calorie intake budget corresponding to the current moment, a dietary intake prompt for the target food in the food image is generated; Dietary intake prompts are provided in the prompting module of wearable devices.

[0009] In one embodiment, before the step of generating a dietary intake suggestion for the target food in the food image based on the total calorie intake and the calorie intake budget corresponding to the current time, the calorie detection method based on the wearable device further includes: Obtain the current eating stage and / or the current user's dietary goals, wherein the dietary goals include at least the weight change-oriented dietary needs; Determine the calorie intake budget for the current moment based on the eating stage and / or the dietary needs.

[0010] In one embodiment, before the step of generating a dietary intake suggestion for the target food in the food image based on the total calorie intake and the calorie intake budget corresponding to the current time, the calorie detection method based on the wearable device further includes: Acquire user motion data and body characteristic parameters collected by other devices connected to the wearable device; The calories burned that day are determined based on the exercise data and the body characteristic data. Based on the calories consumed that day, determine the calorie intake budget for the current moment.

[0011] In one embodiment, after the step of generating a dietary intake suggestion for the target food in the food image based on the total calorie intake and the calorie intake budget corresponding to the current time, the calorie detection method based on the wearable device further includes: If the ratio between the calories consumed that day and the preset calories consumed; If the ratio is outside the preset ratio, the dietary intake prompt is updated based on the calories consumed that day.

[0012] In one embodiment, the prompting module includes at least a bone conduction speaker and a display component, and the step of providing dietary intake prompts in the prompting module of the wearable device includes: The bone conduction speaker outputs a voice broadcast corresponding to the dietary intake prompt; The dietary intake prompts are displayed based on the output of the display component.

[0013] In addition, to achieve the above objectives, this application also proposes a wearable device, the wearable device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the calorie detection method based on the wearable device as described above.

[0014] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the calorie detection method based on wearable devices as described above.

[0015] One or more technical solutions proposed in this application have at least the following technical effects: By combining inertial data from an inertial sensor with data from other sensors, the system determines whether to trigger the calorie recognition process. Once triggered, image recognition is performed. This approach, combining different dimensional parameters, enables seamless triggering of the calorie recognition process, avoiding the hassle of manual operation by the user, improving the accuracy of scene recognition, and reducing false triggers in non-eating scenarios. In portable use scenarios for wearable devices, users can complete calorie detection without complicated operations during daily meals, improving ease of use and data accuracy. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the first embodiment of the calorie detection method based on wearable devices in this application; Figure 2 This is a flowchart illustrating the second embodiment of the calorie detection method based on wearable devices in this application; Figure 3 This is a schematic diagram of the hardware operating environment involved in the calorie detection method based on wearable devices in the embodiments of this application.

[0019] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0021] The main solution of this application embodiment is: to determine whether to trigger the calorie recognition process based on the inertial data of the inertial sensor and the sensor data of other sensors besides the inertial sensor; If so, the current calorie count is determined based on the type and portion size of the food in the food image captured by the image acquisition module.

[0022] In this embodiment, for ease of description, the following description uses a wearable device as the execution subject.

[0023] Because identifying dietary calorie intake using wearable devices requires clicking the calorie recognition function on the device itself, or accessing the calorie recognition function through a connected terminal such as a mobile phone, and then identifying food calories from captured images, current dietary calorie recognition methods are cumbersome.

[0024] Furthermore, exercise and diet data are often scattered across different devices or applications, lacking unified analysis and intelligent linkage.

[0025] Based on this, this application provides a solution that, when performing calorie detection using wearable devices, combines inertial data from an inertial sensor with data from other sensors to determine whether to trigger the calorie recognition process. After triggering the calorie recognition process, image recognition is performed. This approach, by combining different dimensional parameters, achieves seamless triggering of the calorie recognition process, avoiding the tedious manual operation by the user, improving the accuracy of scene recognition, and reducing false triggering in non-eating scenarios. In the portable use scenario of wearable devices, users can complete calorie detection without complicated operations during daily meals, improving ease of use and data accuracy.

[0026] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or wearable device capable of performing the above functions. Wearable devices include wristbands, watches, smart glasses, headphones, and media, etc. The following description uses smart glasses as an example to illustrate this embodiment and the subsequent embodiments.

[0027] In this embodiment, the smart glasses include at least a recognition module, an image acquisition module, an environmental perception module, and a motion sensing module. The motion sensing module is typically an inertial sensor, while the environmental perception module includes at least an ambient light sensor and a distance sensor. The smart glasses can perform multimodal fusion judgment through the ambient light sensor and distance sensor in the environmental perception module, and the accelerometer and gyroscope in the motion sensing module. Specifically, the accelerometer and gyroscope detect the wearer's head posture and activity state, while the data collected by the distance sensor and ambient light sensor are combined to determine whether the user is currently eating. The image acquisition module integrates a high-resolution miniature camera to capture images of food while the user is eating. The recognition module may include a local processing unit and a cloud processing unit, performing calorie detection locally or in the cloud using a food recognition model.

[0028] Furthermore, the smart glasses also include a wireless communication module, a human-computer interaction module, and a power management module. The wireless communication module supports common communication protocols such as Bluetooth, Wi-Fi, and NFC, enabling it to connect with other wearable devices and acquire exercise data or body characteristic parameters such as heart rate, steps, body fat percentage, and sleep quality. The human-computer interaction module may include bone conduction speakers, micro-projectors, or AR (Augmented Reality) display components to provide users with dietary suggestions or exercise reminders. The power management module supports wireless charging and low-power operation strategies.

[0029] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0030] This application provides a calorie detection method based on a wearable device, preferably smart glasses. Please refer to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the calorie detection method based on wearable devices according to this application.

[0031] In this embodiment, the calorie detection method based on wearable devices includes steps S10 to S20: Step S10: Based on the inertial data from the inertial sensor and the sensor data from other sensors besides the inertial sensor, determine whether to trigger the calorie recognition process.

[0032] In this embodiment, the inertial sensors include accelerometers and gyroscopes. Inertial data is the data detected by the accelerometers and gyroscopes, which can capture the posture changes corresponding to actions such as looking down, raising the wrist, and putting food into the mouth during the user's eating process. The calorie recognition process is the process that needs to be triggered when determining the current eating scenario. Other sensors include at least one of ambient light sensors and distance sensors.

[0033] It is understandable that when the wearable device is a head-mounted device such as glasses or headphones, parameters such as the head tilt angle and head angular velocity when the user wears smart glasses or smart headphones can be calculated using inertial data. Optionally, when the wearable device is a hand-mounted device such as a watch, bracelet, or medium, hand posture can be calculated using inertial data. This embodiment and subsequent embodiments use smart glasses as an example for illustration.

[0034] Specifically, when determining whether to trigger the calorie recognition process solely based on inertial data, normal user postures such as simple head turning or looking down can falsely trigger it. Therefore, it is necessary to fuse sensor data from other sensors besides the inertial sensor to determine the calorie recognition process.

[0035] As an optional implementation, based on inertial data, the determination can be made using ambient light sensors and / or distance sensors. During the determination process, data analysis can be performed in a lightweight scene classifier deployed locally on the smart glasses, such as a model built on decision trees or small neural networks. Specifically, after the user puts on the smart glasses, the inertial data collected by the inertial sensors first determines whether the user's head tilt angle and head angular velocity meet the thresholds. If so, it is determined to be an eating state, and a confidence score is output. Simultaneously, distance data collected by the distance sensors determines whether the distance between the glasses and the table is within the eating distance, and the eating state and confidence score are determined based on the eating distance. Similarly, the ambient light intensity is used to determine the eating state and confidence score. Finally, based on a weighted summation of the confidence scores, it is determined whether to trigger the calorie recognition process, where the weights of the summation can be set according to actual needs. In this way, the calorie recognition process is triggered by the multimodal fusion results of inertial data, distance parameters, and / or ambient light data.

[0036] For example, when using inertial sensors, distance sensors, and ambient light sensors for judgment, the inertial sensors of the smart glasses collect user head movement data at a sampling rate of 30Hz. A lightweight decision tree model analyzes the data in real time, extracting the head-down angle and head angular velocity features. When the user maintains a head-down posture while eating, the sensor detects that the head-down angle remains stable at 35° (≥30° threshold) for 3 seconds, and the head angular velocity is 0.08 rad / s (<0.1 rad / s static threshold), perfectly matching the head posture characteristics of eating. The scene classifier classifies this as a high-confidence eating state, assigning a confidence level of 90%. Simultaneously, the distance sensor monitors the real-time distance between the smart glasses and the table or plate. The sensor continuously collects distance values ​​of 32cm, 33cm, and 35cm, all stable within the eating distance range of 20–50cm without significant fluctuations. This indicates that the user's head maintains the stable distance required for eating between the user and the plate. The classifier classifies this as a medium-to-high-confidence eating state, assigning a confidence level of 85%. The ambient light sensor captures the ambient light intensity in real time. The intensity detection values ​​in the ambient light data are 450 lux, 452 lux, and 448 lux, with a fluctuation range of only ±4 lux, which is less than the preset ±10 lux threshold. The stable duration is 4 seconds, which is greater than the 3-second threshold. This meets the lighting characteristics of a stable indoor dining environment. The classifier judges it as a medium-confidence eating state and assigns a confidence level of 80%.

[0037] Finally, when calculating the weights, head posture, as the core eating action indicator, is assigned a weight of 0.5; distance change, as the spatial verification indicator of the eating scene, is assigned a weight of 0.3; and illumination stability, as an auxiliary indicator of the eating environment, is assigned a weight of 0.2. Using the formula "Overall Confidence = Head Posture Confidence × 0.5 + Distance Detection Confidence × 0.3 + Illumination Stability Confidence × 0.2", substituting the above parameters, we get 90% × 0.5 + 85% × 0.3 + 80% × 0.2 = 45% + 25.5% + 16% = 86.5%. The preset calorie recognition process trigger threshold is 80%. Since the overall confidence level is 86.5% ≥ 80%, the smart glasses automatically trigger the calorie recognition process.

[0038] Optionally, in addition to directly using the multimodal fusion results of inertial data, distance parameters, and / or ambient light data to trigger the calorie recognition process, initial verification can be performed first using inertial data. After the initial verification is successful, data can then be acquired using distance sensors and / or ambient light sensors. Therefore, in another feasible implementation, step S10 includes S11~S12: Step S11: If the head posture corresponding to the inertial data meets the preset posture, acquire the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor.

[0039] Step S12: Based on the multimodal fusion results of inertial data, distance parameters, and ambient light data, determine whether to trigger the calorie recognition process.

[0040] In this embodiment, the head posture corresponding to the inertial data is used to determine whether a preset posture is met. If it is, distance parameters collected by the distance sensor and / or ambient light data collected by the ambient light sensor are acquired. Then, multimodal fusion weighted calculation is performed based on the inertial data, ambient light data, and / or distance parameters, and the calorie recognition process is triggered based on the multimodal fusion result. In this way, data is acquired and a secondary judgment is made only after it is determined that the user may be in a eating state, reducing the power consumption of smart glasses.

[0041] The smart glasses automatically collect and analyze data using inertial sensors, ambient light sensors, and distance sensors to determine whether to trigger the calorie recognition process. The entire process requires no manual operation from the user, enabling seamless recognition of eating status and improving ease of use.

[0042] Step S20: If yes, determine the current calories based on the food type and portion size in the food image acquired by the image acquisition module.

[0043] In this embodiment, the image acquisition module is a high-resolution miniature camera integrated into the wearable device. After the calorie recognition process is triggered, the image acquisition module automatically activates the high-resolution miniature camera and captures an image of the food within the current field of view. It can be understood that the food in the image is the food the user is about to consume. The calorie recognition process is triggered each time the user puts food in their mouth.

[0044] As an alternative implementation, after a user wears smart glasses and triggers the calorie recognition process, the captured food image typically needs to be preprocessed first. The preprocessed image is then input into a lightweight food recognition model such as EfficientNet-Lite (a lightweight variant designed by Google for mobile CPU / GPU inference) or a GhostNet variant (a variant of the Ghost module). This model has been trained on a labeled dataset containing tens of thousands of common Chinese and Western foods, supporting simultaneous and parallel recognition of multiple food categories. It can assess food portion size by combining pixel area ratio with prior knowledge of tableware. Simultaneously, it automatically calculates the current calories based on portion size (g) × calories per 100g using a locally cached lightweight calorie database. It should be noted that during image preprocessing, the focus area can be automatically cropped based on detected ROIs (Regions of Interest) for faces / hands, followed by color correction using a standard color chart reference or white balance algorithm, and finally, scale normalization and background suppression are performed using a semantic segmentation model.

[0045] For example, the image recognition model identifies the food in the food image as scrambled eggs with tomatoes and mixed grain rice with a confidence level of 93%. At the same time, it identifies the diameter of the plate as 20cm based on the outline of the plate, calculates that the scrambled eggs with tomatoes account for 40% of the total weight and weighs 120g, and the mixed grain rice accounts for 60% of the total weight and weighs 180g. Then, it retrieves the database data that 100g of scrambled eggs with tomatoes contains 180 kcal and 100g of mixed grain rice contains 130 kcal. Finally, it calculates the current calorie content as 120×1.8+180×1.3=216+234=450 kcal.

[0046] In another alternative implementation, if the confidence level of the output current calories is less than a preset threshold, it can be marked as low confidence and cloud detection can be triggered to transmit the data to the cloud for detection by the cloud model.

[0047] Optionally, when the image acquisition module has multiple cameras, it can acquire food images from different angles using multiple cameras, so that the model can analyze multiple food images from different angles and improve the accuracy of current calorie detection.

[0048] Furthermore, after obtaining the current calorie count of the food to be consumed, this calorie data can be displayed on the smart glasses' display module. For example, the data can be displayed using an AR display component.

[0049] This embodiment provides a calorie detection method based on wearable devices. It fuses inertial data from smart glasses with other multi-dimensional parameters to determine whether to trigger the calorie recognition process. This achieves seamless triggering of the calorie recognition process, avoiding the tedious manual operation by the user, while improving the accuracy of scene recognition and reducing false triggers in non-eating scenarios. In eating scenarios, the collected images are analyzed to identify the food to be eaten, estimate its portion size, and calculate its calories. This completely seamless calorie recognition ensures accuracy and timeliness, while allowing calorie monitoring to be seamlessly integrated into daily eating scenarios, solving the problems of traditional calorie recognition methods that require manual operation, have poor adaptability, and offer a poor user experience.

[0050] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to the first embodiment described above can be referred to the above description and will not be repeated hereafter. Furthermore, before step S11, step S13 is also included: Step S13: Acquire the target inertial data collected by the wristband device connected to the wearable device, and determine the hand posture corresponding to the target inertial data.

[0051] In this embodiment, in addition to automatically collecting and analyzing data through the inertial sensors, ambient light sensors, and distance sensors of the smart glasses, corresponding detection data can also be obtained from associated wearable devices via a wireless communication module. This detection data is then combined to further assess the calorie recognition process, thereby improving the accuracy of the triggered calorie recognition process. The wristband device can be a smart bracelet, smartwatch, or similar device equipped with an inertial sensor.

[0052] Therefore, after determining that the head posture meets the preset posture through inertial data, target inertial data collected by the inertial sensor of the wristband device connected to the smart glasses can also be acquired, thereby determining the hand posture through the target inertial data. For example, the hand posture can be used to determine whether the change of the gesture during a preset time period meets the preset posture corresponding to the action of picking up food with a knife, fork, chopsticks, etc. If the hand posture meets the preset posture, the processing action of step S11 is executed to acquire ambient light data and distance parameters for secondary judgment.

[0053] Furthermore, after step S11, it is possible to determine whether to trigger the calorie recognition process based on the multimodal fusion results of inertial data, target inertial data, distance parameters, and ambient light parameters. That is, based on the original processing method of step S12, the target inertial data collected by the associated device is weighted and processed to improve the accuracy of triggering the calorie recognition process.

[0054] This embodiment provides a calorie detection method based on wearable devices. In addition to detection by smart glasses, it combines the target inertial parameters with a wristband device connected to the smart glasses. The target inertial parameters are then used to make a secondary judgment on the posture detection. Furthermore, the target inertial parameters are used as one of the multimodal fusion data in the calorie recognition process. Through the collaborative action of multiple interconnected devices, the triggering accuracy of non-intrusive calorie recognition is improved.

[0055] Based on the first or second embodiment of this application, the same or similar content as the above embodiments in the third embodiment of this application can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 After step S20, steps S30 to S50 are also included: Step S30: Determine the total calorie intake based on the current calorie intake and the historical calorie intake within the preset time period.

[0056] In this embodiment, after calculating the current calories of the food to be consumed, it is also necessary to provide intake suggestions based on the current calories, thereby providing dietary guidance through smart glasses. When providing dietary guidance, in addition to directly providing intake suggestions based on the current calorie value, it is usually necessary to combine the user's daily calorie budget, or the calorie budget for the current eating stage, such as breakfast or lunch, for analysis. Therefore, the historical calorie intake within a preset time period can be the total intake for the current eating stage, or the total number of calories the user has consumed that day. This data is usually refreshed at a preset time period, such as midnight. Total calorie intake includes the amount already consumed plus the current calories corresponding to the food to be consumed.

[0057] After determining the current calorie intake, the system continuously assesses whether the user's head posture changes during chewing and swallowing based on inertial parameters detected by inertial sensors, thus determining whether the food intake has been completed. Specifically, after each calorie recognition process is triggered and the calorie intake of the current food is determined through multi-dimensional eating detection parameters, it is also necessary to combine subsequent inertial parameters continuously collected by the smart glasses, such as micro-vibrations of the head corresponding to chewing and swallowing, physiological state parameters such as the duration of eating-related physiological fluctuations, and scene contact parameters such as the continuity of tableware contact and the frequency of lip contact, to comprehensively determine whether the user has completed the eating behavior. If the system detects completed food recognition, continuous chewing and swallowing, and physiological parameters maintaining the eating state, it is confirmed that the user has eaten the food, and the current calorie intake is automatically added to the historical calorie intake database for the day or the current meal period. If no subsequent eating behavior loop is detected, such as only triggering recognition but not chewing and swallowing, it means that the food has not been consumed, and the calorie data for this instance is not counted.

[0058] Understandably, recommending foods based solely on the calorie count of a single meal—for example, a food containing 200 calories that is suitable for consumption—does not consider the user's total calorie intake over a preset period, such as the day, the meal, or the weight loss cycle. This could lead to recommendations that conflict with dietary goals. For instance, if a user has already consumed 1800 calories that day, close to their daily goal of 2000 calories, recommending a high-calorie food of 300 calories would result in exceeding the recommended intake. Therefore, it is necessary to calculate the total intake by adding the current intake to the total daily intake, and then make effective recommendations based on this total intake to avoid situations where a single meal is reasonable but the total intake exceeds the recommended amount.

[0059] Step S40: Based on the total calories ingested and the calorie intake budget corresponding to the current moment, generate dietary intake prompts for the target food in the food image.

[0060] In this embodiment, the calorie intake budget corresponding to the current moment can be either the daily intake budget or the intake budget for the current eating stage. Specifically, the timing of the calorie recognition process triggered by the smart glasses determines the current eating stage, and then the preset calorie intake budget corresponding to the eating stage is obtained through a lookup table. It is understood that if the user prioritizes comparing data from the current day, the calorie intake budget will be the total budget for that day; similarly, if the comparison is set to use data from the current eating stage, the budget will be the calculation corresponding to that stage. The calorie intake budget is typically derived from the user's diet plan or directly input by the user on the device.

[0061] After obtaining the total calorie intake and calorie intake budget, the ratio between the total calories consumed that day and the current calorie budget can be calculated, and corresponding dietary intake prompts can be generated based on this ratio. Alternatively, the ratio between the total calories consumed in the current dietary phase and the calorie intake budget for the current dietary phase can be calculated, and corresponding dietary intake prompts can be generated based on this ratio.

[0062] For example, if the total calorie intake is 800 kcal, dividing it by the daily intake budget of 1000 kcal results in a value greater than 0.8. In this case, a message is generated: "Current intake has reached 80% of the daily target; it is recommended to reduce staple foods and increase vegetables." If the calculation is based on the meal stage, and the current meal stage is lunch, with a total calorie intake of 300 kcal and a lunch intake budget of 500 kcal, a message is generated: "Current diet is normal; you can eat normally." If the total calorie intake exceeds the intake budget, a message is generated: "Current intake exceeds the target; it is recommended to replace with low-calorie foods."

[0063] Optionally, the analysis can also be based on the user's actual health status. For example, if the user is at risk of diabetes, a prompt can be generated when a high-sugar beverage is detected: "Sugar-containing beverage detected. It is recommended to replace it with a sugar-free beverage."

[0064] Step S50: Provide dietary intake prompts in the prompt module of the wearable device.

[0065] In this embodiment, the prompt module of the smart earphone can be a bone conduction speaker, an AR display component, etc., to perform voice broadcasting through the bone conduction speaker and to output prompt content through the AR display component.

[0066] This embodiment provides a calorie detection method based on wearable devices. It determines the total calorie intake based on the current calorie intake and historical calorie intake within a preset time period. Then, based on this total calorie intake and the calorie intake budget corresponding to the current moment, it generates a dietary intake prompt for the target food in a food image and outputs the dietary intake prompt in the prompt module of the wearable device. This achieves a closed-loop technology for seamless calorie recognition, avoiding the tedious process of manually summarizing historical intake data and calculating remaining budgets. Furthermore, it combines the calorie data of a single meal with the user's overall dietary plan, ensuring that the dietary intake prompt accurately matches the user's daily dietary goals. This helps users know in real time whether their current food intake meets budget requirements and whether adjustments are needed, solving the problem that traditional calorie recognition can only provide single-data points and cannot offer targeted dietary guidance. The prompt module of the wearable device displays prompt information in real time, ensuring that users can quickly obtain feedback during eating, facilitating timely adjustments to their eating behavior and improving practicality in daily health management scenarios.

[0067] Based on the third example of this application, in the fourth embodiment of this application, the content that is the same as or similar to the third embodiment described above can be referred to the above description, and will not be repeated hereafter. Furthermore, as an optional implementation, after step S40, steps S60-S70 are also included: Step S60: Obtain the current eating stage and / or the current user's dietary goals.

[0068] Dietary goals include at least weight-related dietary needs, such as muscle gain and fat loss, and can also be based on changes in sugar content, such as blood sugar control.

[0069] In this embodiment, in addition to determining the calorie intake budget based on the eating stage, the intake budget can also be determined solely based on the user's dietary goals, or by combining the eating stage and dietary goals. The eating stage can be obtained through current time analysis, while the dietary goal can be a goal set by the user on the application corresponding to the smart glasses. For smart glasses with a simple visual interface, the dietary goal can also be set within that interface.

[0070] Step S70: Determine the calorie intake budget for the current moment based on the eating stage and / or dietary needs.

[0071] As an optional implementation method, the calorie intake budget for the current time period can be determined directly by looking up a table based on the eating stage, such as 500 kcal for lunch and 450 kcal for dinner.

[0072] As another alternative implementation method, when determining the calorie intake budget based on dietary needs, the intake corresponding to different age groups such as normal adults, teenagers, and middle-aged people can also be obtained by referring to a table. For example, if you are trying to lose weight, the calorie budget is 300 kcal, etc. The data corresponding to different age groups are different, which will not be elaborated here.

[0073] As another alternative implementation, when determining the calorie intake budget based on both the eating stage and dietary needs, one can first look up the calorie intake budget for that stage, and then adjust it according to dietary needs. For example, when building muscle, the calorie intake budget could be increased by 100 kcal, while when losing fat, it could be decreased by 100 kcal. Alternatively, a weighted summation method can be used, summing the budgets corresponding to each eating stage and dietary needs to obtain the actual budget.

[0074] Furthermore, if it is necessary to calculate the daily calorie intake budget, the same process shall be followed, and this application will not elaborate further.

[0075] This embodiment provides a calorie detection method based on wearable devices. Based on the eating stage, it combines the user's actual dietary goals, such as muscle gain, fat loss, and blood sugar control, to dynamically calculate the calorie intake budget corresponding to the current moment. By combining multi-source data, the accuracy of the calorie intake budget is improved, thereby improving the accuracy of dietary intake prompts.

[0076] Based on the third example of this application, in the fifth embodiment of this application, the content that is the same as or similar to the third embodiment described above can be referred to the above description, and will not be repeated hereafter. Furthermore, as an optional implementation, after step S40, steps S80 to S100 are also included: Step S80: Obtain the user's motion data and body characteristic parameters collected by other devices connected to the wearable device.

[0077] Other devices include at least wristband devices and body parameter monitoring devices. Wristband devices include smartwatches / bands, while body parameter monitoring devices include smart scales and height / weight measurement devices. These devices are interconnected on the same local area network. Activity data includes the user's cumulative daily steps and data on different activity types such as sitting, walking, and running, which can be used to calculate calorie consumption.

[0078] In this embodiment, a fixed intake budget cannot match an individual's actual daily energy needs. Therefore, by calculating the number of calories consumed by the user on that day, the daily energy consumption level is accurately captured, and the intake budget is dynamically linked to the consumption level. This ensures that the intake budget will not lead to insufficient energy and affect bodily functions due to being lower than the actual consumption, nor will it lead to a calorie surplus and cause abnormal weight problems due to being higher than the actual consumption.

[0079] Specifically, smart glasses can obtain users' daily cumulative steps, sedentary, walking, running and other exercise data from devices such as smartwatches and smart bands, as well as the latest body characteristic parameters such as body fat percentage, basal metabolic rate and muscle mass from devices such as smart scales.

[0080] Step S90: Determine the calories consumed that day based on the exercise data and the body characteristic data.

[0081] After acquiring exercise data and body characteristic data, the data can be aligned with a unified timestamp and then sent to the smart glasses' local energy consumption estimation engine. Combining the metabolic equivalent table with the user's individual parameters, the daily calorie consumption is dynamically calculated.

[0082] For example, the smart bracelet transmits user activity data from 9:00 AM to 6:00 PM, including a cumulative step count of 8200, 6 hours of sitting, 1.5 hours of walking, and 0.5 hours of running. The scale transmits body characteristic parameters of 22% body fat, 1580 kcal / day basal metabolic rate, and 38 kg of muscle mass. The aligned data is sent to the smart glasses' local energy consumption estimation engine. The engine calls the built-in metabolic equivalent table: 1.2 for sitting, 3.3 for walking, and 8.0 for running; combined with the metabolic efficiency coefficient corresponding to the user's muscle mass. The coefficient increases by 0.02 for every 1 kg increase in muscle mass; the adjustment coefficient for body fat percentage is 0.95 for 22% body fat; finally, using the dynamic energy consumption formula: energy consumption for each time period = metabolic equivalent × weight × time × individual coefficient, plus the time-based percentage of basal metabolic rate, the calculated daily calorie expenditure is 2350 kcal.

[0083] Furthermore, in addition to acquiring motion data and body characteristic parameters through interconnected sensors, smart glasses can also collect the user's daily motion data through a built-in motion detection module.

[0084] Step S100: Based on the calories consumed that day, determine the calorie intake budget corresponding to the current moment.

[0085] When determining a calorie intake budget, you can use a table to calculate the calorie intake budget corresponding to the calories already consumed that day, based on the current time. For example, if the calories consumed between 0:00 and 12:00 are x, then the calorie intake budget corresponding to x can be determined by looking up the table as y.

[0086] For example, assuming a user's health goal is to maintain their weight, the calorie intake budget table built into the smart glasses is preset based on individual parameters such as their basal metabolic rate of 1600 kcal / day and muscle mass of 39 kg. At 11:45 on the same day, the local energy consumption estimation engine calculates that 880 kcal have been consumed. By querying the preset table, the calorie intake budget corresponding to 850-900 kcal consumed between 0:00 and 12:00 is 1200 kcal. Therefore, the calorie intake budget corresponding to the current moment is determined to be 1200 kcal.

[0087] In addition, the existing calorie intake budget can be updated based on the ratio between the calories consumed that day and the normal calorie intake. For example, if the calories consumed that day are greater than the normal calorie intake, the calorie intake budget will be increased.

[0088] Optionally, dietary intake suggestions can be updated based on the calories consumed that day. Specifically, the ratio between the calories consumed that day and the preset calorie expenditure is first calculated. If the ratio is outside the preset range, the dietary intake suggestions are updated based on the calories consumed that day. The preset calorie expenditure can be the normal daily calorie expenditure for a normal adult, child, middle-aged person, or elderly person. The daily calorie expenditure is used to determine if the current exercise level is excessive. The preset ratio is typically 0.3 to 1, meaning the daily calorie expenditure is less than the preset calorie expenditure. A ratio outside the preset range indicates either insufficient or excessive exercise. If the ratio is greater than the maximum preset ratio, the user can be advised to "eat more"; if it is less than the minimum preset ratio, the user can be advised to "eat less." For example, if the original dietary intake suggestion was "Current intake has reached 80% of the daily target, you can eat normally," and the user is advised to reduce calorie intake when the calories consumed are low, the dietary intake suggestion could be changed to "Current intake has reached 80% of the daily target, it is recommended to reduce staple foods and increase vegetables." Optionally, the user's sleep quality score, current heart rate, and other parameters can be obtained through watches and wristbands, and these parameters can be used to calculate the calories consumed that day, thereby improving the accuracy of calorie consumption calculation.

[0089] This embodiment provides a calorie detection method based on wearable devices. By integrating multi-dimensional health data from devices such as wristbands, watches, and scales, and calculating or updating the calorie intake budget based on the calories consumed that day, it provides users with personalized dietary intake references that match their real-time consumption status, individual health goals (such as weight maintenance or fat loss), and the current eating time. This allows users to quickly know their reasonable calorie intake without manual calculation or intervention, avoiding calorie surplus or deficiency caused by an imbalance between intake and consumption, improving the scientific and convenient nature of diet management, and building a more comprehensive user profile.

[0090] For example, to aid in understanding the implementation flow of the wearable device-based calorie detection method obtained by combining the above embodiments, this embodiment proposes an optional implementation flow for calorie detection and diet recommendation in the wearable device-based calorie detection method. Specifically: During the eating scene recognition process, multimodal fusion judgment is first performed using the ambient light sensor and distance sensor in the environmental perception module, and the accelerometer and gyroscope in the motion sensing module. When the user's head-down angle continuously exceeds a threshold and the head angular velocity is lower than a set static threshold, it is initially determined that the user may be in a eating state. At the same time, the distance sensor monitors whether the distance between the glasses and the table or plate is stable within the typical eating distance range. In the ambient light data, if the ambient light intensity fluctuation is less than a preset threshold and the duration exceeds 3 seconds, it is considered a stable indoor dining environment. The calorie recognition process is then triggered.

[0091] Once the eating scenario is confirmed, the image acquisition module automatically activates a high-resolution miniature camera to capture images of the food within the current field of view. After preprocessing, the images are input into a lightweight food recognition model to obtain the food type, portion size, and current calorie count. Next, the wireless communication module acquires real-time physiological and activity data from associated wearable devices, including but not limited to current heart rate, daily cumulative steps, activity type (sedentary / walking / running), and sleep quality score from smartwatches / bands; and the latest body fat percentage, basal metabolic rate (BMR), and muscle mass from a smart scale, as well as the user's health profile such as age, gender, height, target weight, and fat loss / muscle gain mode. This data, after being aligned with a unified timestamp, is sent to a local energy consumption estimation engine, which combines the MET (metabolic equivalent) table with individual user parameters to dynamically calculate the daily calorie expenditure (TDEE_est).

[0092] The system then performs calorie estimation and intake assessment. Based on the identified food types and portions, it queries a locally cached lightweight calorie database. If a food does not match the local database or the recognition confidence is low, the image and contextual information (such as geographical location, meal time, and historical preferences) are uploaded to the cloud server via a cloud collaboration module. The cloud then calls a high-precision food recognition model and an expanded calorie database to return accurate calorie and nutritional information. The system combines the local and cloud results to calculate the total calories (kcal_intake) of the current intake and previous intake data, and compares it with the user's daily calorie budget.

[0093] During the comparison process, a lightweight reinforcement learning model is used for analysis. The reward function is the user's long-term health goals of fat loss / sugar control / muscle gain. Contextualized suggestions are generated by combining historical behavior (such as frequently overeating at dinner). If kcal_intake / daily_budget ≥ 0.8 and the daily TDEE_est is low, the suggestion is: current intake has reached 80% of the daily goal; reduce staple foods and increase vegetables. If a high-sugar beverage is detected and the user has a diabetes risk label, a reminder is issued that a sugary beverage has been detected, and a suggestion is made to replace it with a sugar-free beverage. During the interaction, voice announcements are made through bone conduction speakers, and AR display components such as waveguide optical modules are used to overlay text prompts or calorie progress bars at the user's field of vision. Finally, all raw data (after anonymization), recognition results, and user feedback are encrypted and synchronized to the cloud-based health platform. The cloud uses a federated learning framework to aggregate multi-user data, continuously optimizing the food recognition model and personalized recommendation strategy while protecting privacy. Model update packages are regularly pushed to the terminal to achieve continuous evolution of system capabilities.

[0094] This application provides a wearable device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the calorie detection method based on the wearable device in the first embodiment described above.

[0095] The following is for reference. Figure 3 It shows a structural schematic diagram suitable for implementing the wearable device of the embodiments of this application. Figure 3 The wearable device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of this application.

[0096] like Figure 3 As shown, the wearable device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the wearable device. The processing unit 1001, the ROM 1002, and the RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the wearable device to communicate wirelessly or wiredly with other devices to exchange data. While wearable devices with various systems are shown in the figures, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0097] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0098] The wearable device provided in this application employs the calorie detection method based on wearable devices described in the above embodiments, which solves the technical problem of requiring manual operation for calorie identification, which is cumbersome and has low identification efficiency. Compared with the prior art, the beneficial effects of the wearable device provided in this application are the same as those of the calorie detection method based on wearable devices provided in the above embodiments, and other technical features of this wearable device are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0099] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0100] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0101] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the wearable device-based calorie detection method in the above embodiments.

[0102] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory (EPROM, or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, radio frequency (RF), etc., or any suitable combination thereof.

[0103] The aforementioned computer-readable storage medium may be included in the wearable device; or it may exist independently and not assembled into the wearable device.

[0104] The aforementioned computer-readable storage medium carries one or more programs that, when executed by a wearable device, cause the wearable device to: Based on the inertial data from the inertial sensor and the sensor data from other sensors besides the inertial sensor, determine whether to trigger the calorie recognition process; If so, the current calorie count is determined based on the type and portion size of the food in the food image captured by the image acquisition module.

[0105] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0106] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0107] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0108] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the aforementioned calorie detection method based on wearable devices. This solves the technical problem of requiring manual operation for calorie identification, which is cumbersome and inefficient. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the calorie detection method based on wearable devices provided in the above embodiments, and will not be repeated here.

[0109] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for calorie detection based on wearable devices, characterized in that, The calorie detection method based on wearable devices includes: Based on the inertial data from the inertial sensor and the sensor data from other sensors besides the inertial sensor, determine whether to trigger the calorie recognition process; If so, the current calorie count is determined based on the type and portion size of the food in the food image captured by the image acquisition module.

2. The calorie detection method based on wearable devices as described in claim 1, characterized in that, The other sensors include at least a distance sensor and an ambient light sensor. The step of determining whether to trigger the calorie recognition process based on inertial data from the inertial sensor and sensor data from the other sensors besides the inertial sensor includes: If the head posture corresponding to the inertial data meets the preset posture, the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor are obtained. Based on the multimodal fusion results of the inertial data, the distance parameters, and the ambient light data, it is determined whether to trigger the calorie recognition process.

3. The calorie detection method based on wearable devices as described in claim 2, characterized in that, Before the step of acquiring the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor, the calorie detection method based on the wearable device further includes: Acquire target inertial data collected by a wristband device connected to a wearable device, and determine the hand posture corresponding to the target inertial data; If the hand posture meets the preset posture, the steps of obtaining the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor are executed; After the steps of acquiring the distance parameters collected by the distance sensor and the ambient light data collected by the ambient light sensor if the head posture corresponding to the inertial data satisfies the preset posture, the calorie detection method based on wearable devices further includes: Based on the multimodal fusion results of the inertial data, the target inertial data, the distance parameters, and the ambient light data, it is determined whether to trigger the calorie recognition process.

4. The calorie detection method based on wearable devices as described in claim 1, characterized in that, After the step of determining the current calories based on the food type and portion size in the food image acquired by the image acquisition module, the calorie detection method based on wearable devices further includes: Based on the current calories and the historical calories consumed within the preset time period, the total calories consumed are determined; Based on the total calories ingested and the calorie intake budget corresponding to the current moment, a dietary intake prompt for the target food in the food image is generated; Dietary intake prompts are provided in the prompting module of wearable devices.

5. The calorie detection method based on a wearable device as described in claim 4, characterized in that, Before the step of generating a dietary intake suggestion for the target food in the food image based on the total calorie intake and the calorie intake budget corresponding to the current moment, the calorie detection method based on the wearable device further includes: Obtain the current eating stage and / or the current user's dietary goals, wherein the dietary goals include at least the weight change-oriented dietary needs; Determine the calorie intake budget for the current moment based on the eating stage and / or the dietary needs.

6. The calorie detection method based on a wearable device as described in claim 4, characterized in that, Before the step of generating a dietary intake suggestion for the target food in the food image based on the total calorie intake and the calorie intake budget corresponding to the current moment, the calorie detection method based on the wearable device further includes: Acquire user motion data and body characteristic parameters collected by other devices connected to the wearable device; The calories burned that day are determined based on the exercise data and the body characteristic data. Based on the calories consumed that day, determine the calorie intake budget for the current moment.

7. The calorie detection method based on a wearable device as described in claim 6, characterized in that, After the step of generating dietary intake prompts for the target food in the food image based on the total calorie intake and the calorie intake budget corresponding to the current moment, the calorie detection method based on wearable devices further includes: If the ratio between the calories consumed that day and the preset calories consumed; If the ratio is outside the preset ratio, the dietary intake prompt is updated based on the calories consumed that day.

8. The calorie detection method based on a wearable device as described in claim 4, characterized in that, The prompting module includes at least a bone conduction speaker and a display component, and the step of providing dietary intake prompts in the prompting module of the wearable device includes: The bone conduction speaker outputs a voice broadcast corresponding to the dietary intake prompt; The dietary intake prompts are displayed based on the output of the display component.

9. A wearable device, characterized in that, The wearable device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the calorie detection method based on the wearable device as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the calorie detection method based on a wearable device as described in any one of claims 1 to 8.

Images