34results about How to "Improve step counting accuracy" patented technology

The invention provides a step counting method based on the walking state, comprising the following steps: continuously acquiring acceleration data from an acceleration sensor; calculating the global acceleration; and then calculating the linear acceleration ACC on three axes x , ACC y and ACC z ; Judgment ACC x , ACC y and ACC z Whether it exceeds the set threshold, if it exceeds the set threshold, mark this set of data as abnormal data and clear it; calculate the component of linear acceleration in the direction of gravity to obtain ACC g ; set ACC g drawn as the first curve, the ACC y Draw the second curve, and overlap the first curve and the second curve to form an overlapping image; set the walking identification condition according to the walking characteristics, use the walking identification condition to analyze the overlapping image, and judge whether walking occurs; when it is judged that walking occurs, the walking ACC at time g The value calculates the step length and calculates the stride frequency to obtain walking data. Compared with the prior art, the present invention can eliminate most equipment shaking interference, provide step length and step frequency data, and effectively improve step counting accuracy.

The invention discloses a novel step counting method based on an accelerometer. The novel step counting method comprises following steps of first, pre-processing acceleration data; second, acquiring azero set S1 of the acceleration data according to a zero crossing detection method; third, acquiring a prediction range of a next peak-to-valley value according to period constraint; fourth, refiningthe prediction range obtained in the third step according to dynamic time warping constraint; fifth, searching a maximum or a minimum, a next peak or valley, within the prediction range obtained in the fourth step; sixth, converting detection results of the peak and valley into a detection result of zero point, and acquiring a predicted zero point; and seventh, eliminating fake zero points in S1according to the predicted zero point, and acquiring a real zero set S2, wherein every two zero points in S2 are recorded as one step. The method solves a problem that a pedestrian counts steps in different motion modes, has high practicability in real life and has a bright application prospect.

The invention belongs to the technical field of communication and provides a step counting method, a step counting device and a mobile terminal. The step counting method comprises the following steps: acquiring accelerated speed information collected by a sensor in the mobile terminal, and establishing a corresponding accelerated speed changing curve graph according to the accelerated speed information; acquiring a peak value of the accelerated speed and the frequency corresponding to the peak value of the accelerated speed according to the accelerated speed changing curve graph; and when the peak value of the accelerated speed is larger than a preset accelerated speed threshold and the frequency corresponding to the peak value of the accelerated speed is smaller than a preset frequency threshold, determining that a user holding the mobile terminal moves one step, and adding 1 to the counted numerical value. By virtue of the step counting method, the false step counting can be effectively prevented, and the step counting precision can be improved.

The invention provides a step counting method and a terminal device. The step counting method comprises the steps: acquiring an acceleration magnitude of the terminal device in real time; determining maximal values and minimal values in the acceleration magnitude; transforming the frequency domain of the acceleration magnitude between each two adjacent maximal values and two adjacent minimal values when the difference between the adjacent maximal value and minimal value is greater than an acceleration threshold value, and acquiring a walking frequency in the time corresponding to the two adjacent maximal values or the two adjacent minimal values; determining that a user carrying the terminal device moves a step in the corresponding time if the walking frequency in the corresponding time is greater than a preset frequency threshold value, and adding 1 to the step number. By adopting the technical scheme, the probability in misjudging when the step is counted can be reduced, and the step counting precision is improved.

The invention provides a step calculation method, which reduces the interference of body situation of a user in step calculation and improve the step calculation accuracy. The method comprises the following steps: based on a step calculation device with sound recognition function, starting walking, prompting the currently various walking states of the step calculation device, and collecting the user step parameters by a sensor; setting an acceleration threshold for step calculation; collecting the acceleration parameter during walking through the sensor, and the like. Information of walking status corresponding steps can be obtained by combining the classification of acceleration values according to the walking states and the check and countering of an envelope line, the user experience is increased, and the step calculation accuracy of the user with weak or bad body status can be improved.

The present application provides a method and device for counting steps based on a filter bank, which includes: after acquiring a plurality of first motion signals of the monitored user, respectively performing denoising processing on the plurality of first motion signals to obtain a plurality of second motion signals. Motion signal, and each second motion signal is processed based on the preset filter bank to obtain the characteristic signal corresponding to each second motion signal, and the target characteristic signal is obtained according to the characteristic signal corresponding to each second motion signal, according to The target feature signal counts the number of steps of the monitored user, and processes the second motion signal through each set of filter banks in the preset filter bank to obtain the feature signal after certain processing in different directions, and suppress the impact of noise on the feature influence of the signal. And a target feature signal is determined according to the feature signal corresponding to each second motion signal, so that the target feature signal for counting can be determined from multiple feature signals, so as to improve the step counting accuracy.

The invention discloses a step counting method, a step counting device and a computer-readable storage medium. The step counting method includes: acquiring acceleration data, the acceleration data including effective peaks; determining the difference between the current effective peak and the last acquired effective peak. time interval between; when the time interval is greater than the preset interval, obtain the data area of ​​the effective peak until the next time interval is greater than the preset interval; when the number of effective peaks in the data area is less than the When the preset value is determined, the feature information of the effective peaks in the data area is determined; when the feature information of the effective peaks in the data area meets the preset condition, the step counting data is updated. The invention provides a step counting method, a step counting device and a computer-readable storage medium, aiming at solving the problem of low step counting accuracy of the step counting device in the prior art.

The invention provides a high-precision and high pseudo-step removing human step-counting method integrating power consumption management. According to the method, a triaxial linear acceleration sensor is used for acquiring original waveforms, sampling time intervals of the original waveforms are adaptively switched according to the change of the original waveforms, the original waveforms of axes are subjected to high-pass filtering and absolute value mode solving one by one, the triaxial waveforms are summed, subjected to low-pass filtering and then respectively smoothened by three points and fourteen points, the waveforms smoothened by the fourteen points serve as dynamic threshold waveforms, the waveforms smoothened by the three points are delayed by eight points and serve as step-counting waveforms, finally, peak values of the step-counting waveforms and the dynamic threshold waveforms are compared, and the time interval between two step-counting waveforms is judged to determine whether the step-counting waveforms are normal waveforms or not. By the step-counting method, pseudo-steps can be effectively removed, step-counting precision is improved, and power consumption is effectively managed.

The invention discloses a step calculation method. The step calculation method comprises following steps: after acquisition of three axises acceleration of arms and hands swinging by a pedometer, reference acceleration of the three axises acceleration in a preset acquisition time is subjected to smoothing so as to obtain processed acceleration; effective characteristic acceleration is obtained from the processed acceleration, and time position of the effective characteristic acceleration in the preset acquisition time is recorded, and a first time window taking the time position of an object characteristic acceleration in the effective characteristic acceleration as the center is determined; it is determined that whether the processed acceleration possesses periodicity based on similarity of processed acceleration waveform in the first time window with processed acceleration waveform in a second time window; if the processed acceleration possesses periodicity, the time length of the first time window is determined to be a step period, and step calculation of the processed acceleration is carried out according to the step period. The step calculation method is capable of reducing miscounting or missing counting in step calculation, and increasing accuracy of step calculation.

The invention discloses a step counting method and device, which are used to solve the problem of low step counting precision in the prior art. The step counting device collects the information of the wearing limbs, and according to the wearing limb information, determines the limb type wearing the step counting device, and the limb type includes dominant limbs or non-dominant limbs; the step counting device collects and wears the step counting device The step counting information of the limbs in a step counting cycle, in the corresponding relationship between the limb type and the step counting standard reference information, find the step counting standard reference corresponding to the limb wearing the step counting device in this step counting cycle Information, according to the step counting information and the found step counting standard reference information, determine whether the user is in the walking state in the current step counting cycle; when it is determined that the user is in the walking state in the current step counting cycle, then Calculate the number of steps according to the step counting information. By adopting the scheme, the step counting error is reduced, and the step counting precision is improved.

The invention discloses a pedometer. The pedometer comprises: a gathering module for gathering the triaxial acceleration of swinging of an arm according to a preset sampling frequency; a processing module for carrying out smoothing processing on the reference acceleration of the triaxial acceleration within a preset acquisition duration to obtain a processed acceleration; a first acquisition module for acquiring an effective characteristic acceleration from the processed acceleration; a recording module for recording the time position of the effective characteristic acceleration; a determination module for determining a first time window adopting the time position of a target characteristic acceleration as a center; a calculating module for calculating the similarity of processed acceleration wave shapes in the first time window and a second time window; a first judging module for judging whether the acceleration is periodic or not according to the similarity; and a step counting module for determining the duration of the first time window as a gait period when the acceleration is periodic and carrying out step counting on the processed acceleration according to the gait period in order to obtain effective steps. The pedometer can reduce the false counting or miss counting in the step counting process, and improves the step counting accuracy.