A method for user identification and abnormal impedance filtering of an 8-electrode body fat scale

Abstract

The present invention discloses a method for user identification and abnormal impedance filtering of an 8-electrode body fat scale, comprising the following steps: information entry, wherein the subject enters height, age, and gender under a blank user ID; the body fat scale records the most recently tested weight and impedance data of each ID, collectively referred to as "user identification features"; the cosine similarity of the impedance is measured and grouped for judgment; this method can greatly improve the recognition accuracy of user identification when multiple users have similar weights and filter out some impedance anomalies that are not easy to detect.

Description

Technical Field

[0001] The present invention relates to the technical field of body fat scales, and in particular to a method for user identification and abnormal impedance filtering of an 8-electrode body fat scale. Background Art

[0002] Body fat scales can measure body fat percentage. Before measuring body fat percentage, the user's information, such as height, gender, and age, is entered. This information, combined with the weight and bioelectrical impedance measured by the scale, is then substituted into a specified formula to determine the subject's body fat percentage. Most current body fat scales use a pre-entered method, requiring users to select their parameters from a pre-set user list before starting the measurement. However, users sometimes forget to select their parameters and simply step onto the scale. Therefore, some scales have added a post-entry feature, prompting users to select their user information after their weight and bioelectrical impedance are measured. Because body fat scales support multiple users, user information is recorded separately by user ID, allowing users to select their desired user ID when using the scale. To enhance the user experience, body fat scales incorporate a user identification feature. Based on the subject's current weight, the scale searches the user list for the user ID whose last measured weight is closest to their current weight and recommends it to the subject for confirmation. Once confirmed, body fat calculations can begin. However, if there are two or more users in the user list whose recorded weights are similar, the performance of the user identification function will not be very good, and there is a high possibility of incorrect recommendations.

[0003] The bioelectrical impedance measurement of body fat scales is easily affected by the measurement posture of the subject and the contact between the subject and the measuring electrodes, resulting in measurement anomalies. General body fat scales can only identify some obvious impedance measurement errors. For example, if the subject does not step on the scale barefoot, the abnormal impedance will be used to calculate the body fat percentage, which may result in some outrageous results, greatly affecting the subject's satisfaction with the product. Summary of the Invention

[0004] In view of the problems existing in the prior art, the purpose of the present invention is to provide a method for user identification and abnormal impedance filtering of an 8-electrode body fat scale.

[0005] To solve the above problems, the present invention adopts the following technical solutions.

[0006] A method for user identification and abnormal impedance filtering of an 8-electrode body fat scale, step 1: information entry, the subject enters height, age, and gender under the blank user ID;

[0007] Step 2: The body fat scale will record the most recently tested weight and impedance data for each ID. Assuming single-frequency measurement is used, the recorded data includes: W weight, LA left hand impedance, RA right hand impedance, TR trunk impedance, LL left foot impedance, RL right foot impedance, collectively referred to as "user identification features";

[0008] Step 3: Calculate the average value of each parameter of the recent "user identification feature", recorded as: Wm, LAm, RAm, TRm, LLm, RLm;

[0009] Step 4: The subject is measured on the scale without selecting a user ID. The latest "user identification features" are measured, recorded as: Wn, LAn, RAn, TRn, LLn, RLn. The user ID list is traversed and the cosine similarity between the vectors [Wm, LAn, RAn, TRm, LLm, RLm] and [Wn, LAn, RAn, TRn, LLn, RLn] is calculated.

[0010] If the subject selects a user ID, including the selection before the test and the confirmation of the selection after automatic identification, the cosine similarity of the impedance measured this time and the impedance measured most recently saved by the user ID is calculated, and grouping is performed based on the cosine similarity.

[0011] Cosine similarity ≥ 0.97, impedance is normal;

[0012] If the cosine similarity is less than 0.93 and less than 0.97, the user is asked whether to retake the test;

[0013] If the cosine similarity is ≤ 0.93, retest is mandatory;

[0014] If the impedance is normal or the user chooses not to retest, the body fat percentage can be calculated using the impedance of this test and the impedance information can be saved.

[0015] After the user retests, the cosine similarity will be calculated twice.

[0016] The cosine similarity between the re-measured impedance and the most recently measured impedance saved by the user ID,

[0017] The cosine similarity between the remeasured impedance and the impedance before remeasurement,

[0018] If one of these two calculations meets the normal impedance requirement, the current impedance value is used to calculate the body fat percentage and the impedance information is saved. Otherwise, the user is forced to re-measure. The above process filters out some impedances that have low cosine similarity with historical records, reducing the probability of abnormal results.

[0019] As a further improvement to the present invention, between steps 1 and 2, the first time a user uses the scale with a specific ID, they will be prompted to complete two consecutive instructional tests. These tests will only record their weight and impedance information, not their body fat percentage. After the user completes both instructional tests, the user will be notified that the instructional phase is complete and can use the scale normally.

[0020] As a further improvement to the present invention, each valid user ID is assigned a cosine similarity result for its "user identification feature." The cosine similarities calculated for all user IDs are ranked, and the user ID with the highest cosine similarity is recommended to the test subject for confirmation. After the test subject confirms, the user information for that user ID is extracted, and the user information, along with the weight and impedance obtained from the test, is substituted into a formula to calculate body fat percentage.

[0021] If the subject does not confirm the user ID with the highest cosine similarity recommended by the body fat scale, the body fat scale will recommend the user ID with the second highest cosine similarity to the subject. If the subject confirms that the user ID with the second highest cosine similarity matches their own, the body fat scale will calculate their body fat. Otherwise, the subject will be prompted to whether to create a new user or continue to select from the user ID list;

[0022] If the subject chooses to create a new user, the body fat scale will jump to the user information entry process. If the subject chooses to continue selecting from the user ID list, the body fat scale will be recommended to the user in descending order of cosine similarity for confirmation;

[0023] After the subject confirms the user ID, the newly measured "user identification feature" data will be saved under the ID name for use in subsequent user identification.

[0024] Through the above process, even if the subject does not select a user ID, the user parameters with the highest matching degree are quickly found for the user. Because impedance information and weight are added to calculate cosine similarity, the probability of different subjects identifying the same user ID is greatly reduced, which greatly improves the user experience.

[0025] As a further improvement of the present invention, the 5-segment impedance model is: left hand, right hand, trunk, left foot, and right foot;

[0026] When using, the left hand, right hand, left foot, and right foot will be connected to a current electrode and a voltage electrode body fat scale, adding up to a total of 8 electrodes;

[0027] When measuring, the body fat scale applies a fixed-frequency AC signal between two current electrodes, then measures the voltage drop between two voltage electrodes, and then calculates the impedance between the two points based on Ohm's law, as shown below:

[0028] Impedance between hands: Z12 = Z1 + Z2;

[0029] Impedance between the two feet: Z34 = Z3 + Z4;

[0030] Impedance from left hand to left foot: Z13 = Z1 + Z5 + Z3;

[0031] Impedance from left hand to right foot: Z14 ​​= Z1 + Z5 + Z4;

[0032] Impedance from right hand to left foot: Z23 = Z2 + Z5 + Z3;

[0033] Impedance from right hand to right foot: Z24 = Z2 + Z5 + Z4;

[0034] Through mathematical calculations we can get:

[0035] Left-hand impedance: Z1 = (Z14 + Z12 - Z24) / 2;

[0036] Right-hand impedance: Z2 = (Z14 + Z12 - Z24) / 2;

[0037] Left foot impedance: Z3 = (Z34 + Z23 - Z24) / 2;

[0038] Right foot impedance: Z4 = (Z34 - Z23 + Z24) / 2;

[0039] Trunk impedance: Z5 = (Z13 + Z14 + Z23 + Z24 - 2*Z34 - 2*Z12) / 4.

[0040] Beneficial effects of the present invention

[0041] Compared with the prior art, the advantages of the present invention are:

[0042] The present invention proposes a user identification and abnormal impedance filtering method for use on an 8-electrode body fat scale, which can greatly improve the recognition accuracy of users when multiple users have similar weights and filter out some impedance anomalies that are not easy to find. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 Schematic diagram of the impedance structure of the human body 5-segment model of the present invention. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention; it is obvious that the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0045] See also Figure 1 The principle of the 8-electrode body fat scale is to regard the human body as a 5-segment impedance model, namely: left hand, right hand, torso, left foot, and right foot.

[0046] When in use, the left hand, right hand, left foot, and right foot will be connected to a current electrode and a voltage electrode body fat scale. There are a total of 8 electrodes, so it is called an 8-electrode body fat scale.

[0047] When measuring, the body fat scale applies a fixed frequency AC signal between two current electrodes, then measures the voltage drop between two voltage electrodes, and then calculates the impedance between the two points according to Ohm's law, for example

[0048] Impedance between hands: Z12 = Z1 + Z2

[0049] Impedance between the two feet: Z34=Z3+Z4

[0050] Impedance from left hand to left foot: Z13 = Z1 + Z5 + Z3

[0051] Impedance from left hand to right foot: Z14 ​​= Z1 + Z5 + Z4

[0052] Impedance from right hand to left foot: Z23=Z2+Z5+Z3

[0053] Impedance from right hand to right foot: Z24=Z2+Z5+Z4.

[0054] Through mathematical calculations we can get:

[0055] Left-hand impedance: Z1 = (Z14 + Z12 - Z24) / 2

[0056] Right-hand impedance: Z2 = (Z14 + Z12 - Z24) / 2

[0057] Left foot impedance: Z3 = (Z34 + Z23 - Z24) / 2

[0058] Right foot impedance: Z4 = (Z34 - Z23 + Z24) / 2

[0059] Trunk impedance: Z5 = (Z13 + Z14 + Z23 + Z24 -2*Z34 -2*Z12) / 4

[0060] After obtaining the impedance of the 5-segment model of the human body, plus information such as height, weight, age, and gender, you can substitute it into the formula to calculate the body fat percentage of the human body.

[0061] To obtain a more accurate body fat percentage, 8-electrode body fat scales often use multi-frequency sampling. This means using AC signals of varying frequencies to measure the body's impedance. The impedance measured at different frequencies varies; generally speaking, the higher the frequency, the lower the measured impedance. Using more impedance information, the body fat percentage formula developed will be more accurate.

[0062] In the user identification function, including information entry, the subject enters height, age, and gender under the blank user ID;

[0063] The first time a user uses an ID, they will be asked to complete two consecutive instructional tests. These tests will only record their weight and impedance information, not their body fat percentage. Once the user completes both instructional tests, the user will be notified that the instructional phase is over and can use the scale normally.

[0064] The body fat scale will record the weight and impedance data of the two most recent tests for each ID. Assuming single-frequency measurement is used, the recorded data includes: W (weight), LA (left hand impedance), RA (right hand impedance), TR (trunk impedance), LL (left foot impedance), RL (right foot impedance), collectively referred to as "user identification features."

[0065] Calculate the average value of each parameter of the two most recent "user identification features", recorded as: Wm, LAm, RAm, TRm, LLm, RLm.

[0066] The subject is measured on the scale without selecting a user ID. The latest "user identification feature" is measured, recorded as: Wn, LAn, RAn, TRn, LLn, RLn. The user ID list is traversed, and the cosine similarity between the vectors [Wm, LAn, RAn, TRm, LLm, RLm] and [Wn, LAn, RAn, TRn, LLn, RLn] is calculated. Each valid user ID receives a cosine similarity result for its "user identification feature." The cosine similarities calculated for all user IDs are ranked, and the user ID with the highest cosine similarity is recommended to the subject for confirmation. After the subject confirms, the user ID's user information is extracted and substituted into the formula to calculate body fat percentage using this information, along with the weight and impedance obtained from the test.

[0067] If the subject does not confirm the user ID with the highest cosine similarity recommended by the body fat scale, the body fat scale will recommend the user ID with the second highest cosine similarity to the subject. If the subject confirms that the user ID with the second highest cosine similarity matches his or her own, the body fat scale will calculate body fat. Otherwise, the subject will be prompted whether to create a new user or continue to select from the user ID list.

[0068] If the subject chooses to create a new user, the body fat scale will jump to the user information entry process. If the subject chooses to continue selecting from the user ID list, the body fat scale will recommend to the user for confirmation in descending order of cosine similarity.

[0069] In the abnormal impedance filtering step, after the subject confirms the user ID, the newly measured "user identification feature" data will be saved under the ID name for use in subsequent user identification.

[0070] Through the above process, even if the subject does not select a user ID, the user parameters with the highest matching degree can be quickly found for the user. Because impedance information and weight are added to calculate cosine similarity, the probability of different subjects identifying the same user ID is greatly reduced, which greatly improves the user experience.

[0071] The subjects selected their user ID, including selection before the test and confirmation of the selection by the subjects after automatic identification.

[0072] Calculate the cosine similarity between the impedance measured this time and the most recently measured impedance saved by the user ID

[0073] Grouping by cosine similarity

[0074] Cosine similarity ≥ 0.97, impedance is normal

[0075] 0.93 < cosine similarity < 0.97, ask the user whether to retest

[0076] Cosine similarity ≤ 0.93, forced retest

[0077] If the impedance is normal or the user chooses not to retest, the impedance of this test can be used to calculate the body fat percentage and save the impedance information.

[0078] After the user retests, the cosine similarity will be calculated twice.

[0079] Cosine similarity between the remeasured impedance and the most recently measured impedance saved by the user ID

[0080] Cosine similarity of the remeasured impedance and the impedance before remeasurement

[0081] If one of these two calculations meets the normal impedance requirement (cosine similarity ≥ 0.97), the current impedance value will be used to calculate the body fat percentage and the impedance information will be saved. Otherwise, the user will be forced to re-measure.

[0082] The above process filters out some impedances that have low cosine similarity with historical records, reducing the probability of abnormal measurement results.

[0083] The above user identification and impedance filtering are explained using single-frequency measurement as an example. If multi-frequency measurement technology is currently used, impedance values ​​at multiple frequencies will be measured. By adding the impedances at multiple frequencies to the cosine similarity count, it can be used in multi-frequency measurement.

[0084] Multi-frequency measurement can also provide an additional judgment of impedance abnormality. If the high-frequency impedance of a segment is lower than the low-frequency impedance, the user can be reminded to re-measure.

[0085] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto. Any person skilled in the art who, within the technical scope disclosed by the present invention, makes equivalent substitutions or modifications based on the technical solutions and improved concepts of the present invention shall be covered by the scope of protection of the present invention.

Claims

1. A method for user identification and abnormal impedance filtering of an 8-electrode body fat scale, characterized in that: The following steps are involved: Step 1: Information entry: the subject enters height, age, and gender under the blank user ID; Step 2: The scale will record the most recent weight and impedance data for each ID. Assuming single-frequency measurement is used, the recorded data includes: W weight, LA left arm impedance, RA right arm impedance, TR trunk impedance, LL left foot impedance, RL right foot impedance, collectively referred to as "user identification features"; Step 3: Calculate the average value of each parameter of the most recent "user identification feature", recorded as: Wm, LAm, RAm, TRm, LLm, RLm; Step 4: The subject is weighed without selecting a user ID. The latest "user identification features" are measured, recorded as: Wn, LAn, RAn, TRn, LLn, RLn. The user ID list is traversed, and the cosine similarity between the vectors [Wm, LAn, RAn, TRm, LLm, RLm] and [Wn, LAn, RAn, TRn, LLn, RLn] is calculated. Each valid user ID will receive a cosine similarity result for its "user identification feature." The cosine similarities calculated for all user IDs are then ranked, and the user ID with the highest cosine similarity is recommended to the test subject for confirmation. After the test subject confirms, the user ID's user information is extracted and substituted into the formula to calculate body fat percentage using the user information, weight, and impedance obtained from the test. If the subject does not confirm the user ID with the highest cosine similarity recommended by the body fat scale, the body fat scale will recommend the user ID with the second highest cosine similarity to the subject. If the subject confirms that the user ID with the second highest cosine similarity matches their own, the body fat scale will calculate their body fat. Otherwise, the subject will be prompted to whether to create a new user or continue to select from the user ID list; If the subject chooses to create a new user, the body fat scale will jump to the user information entry process. If the subject chooses to continue selecting from the user ID list, the body fat scale will be recommended to the user in descending order of cosine similarity for confirmation; After the subject confirms the user ID, the newly measured "user identification feature" data will be saved under the ID name for use in subsequent user identification; If the subject selects a user ID, including the selection before the test and the confirmation of the selection after automatic identification, the cosine similarity of the impedance measured this time and the impedance measured most recently saved with the user ID is calculated, and grouping is performed based on the cosine similarity: Cosine similarity ≥ 0.97, impedance is normal; If the cosine similarity is less than 0.93 and the user is asked whether to retake the test, Cosine similarity ≤ 0.93, forced retest; If the impedance is normal or the user chooses not to retest, the body fat percentage can be calculated using the impedance of this test and the impedance information can be saved. After the user retests, the cosine similarity will be calculated twice. The cosine similarity between the re-measured impedance and the most recently measured impedance saved by the user ID, The cosine similarity between the remeasured impedance and the impedance before remeasurement, If one of these two calculations meets the normal impedance requirement, the current impedance value will be used to calculate the body fat percentage and the impedance information will be saved. Otherwise, the user will be forced to re-measure.

2. The method for user identification and abnormal impedance filtering of an 8-electrode body fat scale according to claim 1, characterized in that: Between step one and step two, when a certain ID is used for the first time, the subject will be required to perform two consecutive teaching tests according to the prompts. These two tests will only record the subject's weight and impedance information, and will not display the body fat percentage. After the user completes the two teaching tests, the user will be prompted that the teaching phase is over and can use the body fat scale normally.

3. The method for user identification and abnormal impedance filtering of an 8-electrode body fat scale according to claim 1, characterized in that: The 5-segment impedance model is: left hand, right hand, trunk, left foot, and right foot; When using, the left hand, right hand, left foot, and right foot will be connected to a current electrode and a voltage electrode body fat scale, adding up to a total of 8 electrodes; When measuring, the body fat scale applies a fixed-frequency AC signal between two current electrodes, then measures the voltage drop between two voltage electrodes, and then calculates the impedance between the two points based on Ohm's law, as shown below: Impedance between hands: Z12= Z1+Z2; Impedance between the two feet: Z34= Z3+Z4; Impedance from left hand to left foot: Z13= Z1+Z5+Z3; Impedance from left hand to right foot: Z14= Z1+Z5+Z4; Impedance from right hand to left foot: Z23= Z2+Z5+Z3; Impedance from right hand to right foot: Z24= Z2+Z5+Z4; Through mathematical calculations we can get: Left-hand impedance: Z1 = (Z14 + Z12 - Z24) / 2; Right-hand impedance: Z2 = (Z14 + Z12 - Z24) / 2; Left foot impedance: Z3= (Z34 + Z23 - Z24) / 2; Right foot impedance: Z4 = (Z34 - Z23 + Z24) / 2; Trunk impedance: Z5 = (Z13 + Z14 + Z23 + Z24 - 2* Z34 - 2* Z12) / 4.

Images