Doppler-based bathroom fall detection method, apparatus, and radar

Abstract

The application provides a Doppler-based bathroom fall detection method and device and radar, the installation height of the radar is not lower than the installation height of the shower, and the method comprises the following steps: acquiring the track of a target, and acquiring the associated point cloud associated with the track of the target frame by frame, and acquiring the micro-Doppler array of each frame of associated point cloud; determining whether it is a showering scene according to the micro-Doppler array of a continuous first preset number of frames; in the showering scene, acquiring the micro-Doppler array of a continuous second preset number of frames, and calculating the energy Doppler distribution similarity of the associated point cloud of the corresponding second preset number of frames according to the micro-Doppler array of the continuous second preset number of frames; determining the target state according to the energy Doppler distribution similarity of the associated point cloud of the second preset number of frames; and if the target changes from an active state to an inactive state, it is determined that the target falls. The application can improve the target fall discrimination accuracy in the showering scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar detection, in particular to a Doppler-based bathroom fall detection method and device and radar. BACKGROUND

[0002] With the aging phenomenon, the safety monitoring of the elderly is concerned. Among the potential risk factors in the life scene of the elderly, accidental falls account for a high proportion. Among the many fall cases, bathroom falls account for a high proportion.

[0003] At present, there are several types of fall detection methods: the first type is a video processing scheme based on visible light, which obtains a human behavior sequence through an optical sensor and judges a fall event by analyzing the image. Due to privacy issues, the video scheme is not suitable for fall detection in the shower scene; the second type is a fall detection based on sound signals. Due to water interference during showering, the fall detection scheme based on sound signals is also not suitable for the shower scene; the third type is a wearable device, which obtains posture and position information through a miniature sensor such as an acceleration sensor and a gyroscope, and judges whether to fall after processing the information. In the shower scene, the wearable device is not easy to accept; the fourth type is based on environmental perception, which mainly uses sensors placed in the bathroom to capture the influence of human behavior on signals and analyze signal echo data to make a fall judgment. This method has the characteristics of non-contact, not affecting normal life, high privacy, etc. Related environmental perception schemes such as ultrasonic sensors have a high false negative rate and low reliability; infrared devices have low reliability due to water interference, and some infrared devices also have the risk of privacy leakage; fall detection based on millimeter wave radar is a non-contact detection method, which has strong applicability, does not invade privacy, and can be continuously monitored for a long time.

[0004] However, the current radar-based fall detection is mainly based on a sample set to train a model. Due to the fusion of people and water point clouds in the shower scene, the signal of the person is submerged, resulting in low detection accuracy. SUMMARY

[0005] Therefore, the present application provides a Doppler-based bathroom fall detection method, device, radar and storage medium, which can solve the problem of low fall discrimination accuracy in the shower scene.

[0006] In a first aspect, the embodiments of the present application provide a Doppler-based bathroom fall detection method, which is applied to a radar, the installation height of the radar is not lower than the installation height of the shower head, and the method comprises:

[0007] Obtaining a track of a target and obtaining an associated point cloud associated with the track of the target frame by frame;

[0008] For each frame of associated point cloud, obtain the micro-Doppler array of the associated point cloud of that frame. The micro-Doppler array includes the energy distribution of the associated point cloud of that frame in multiple consecutive but non-overlapping Doppler intervals. The Doppler intervals include positive velocity intervals and negative velocity intervals.

[0009] Based on the micro-Doppler array of a consecutive first preset number of frames, determine whether it is a shower scene;

[0010] In the shower scenario, obtain the micro-Doppler array of a consecutive second preset number of frames, and calculate the energy Doppler distribution similarity of the associated point cloud of the corresponding second preset number of frames based on the micro-Doppler array of the consecutive second preset number of frames.

[0011] The target state is determined based on the energy Doppler distribution similarity of the associated point cloud of the second preset frame number. The target state includes whether the target is in an active state or the target is in an inactive state.

[0012] If the target changes from an active state to an inactive state, then the target is determined to have fallen.

[0013] In one possible implementation, obtaining the micro-Doppler array of the associated point cloud for each frame includes:

[0014] For any point cloud in the associated point cloud of this frame, calculate the Doppler interval to which the point cloud belongs based on the radial velocity of the point cloud;

[0015] For each Doppler interval, the energy value of the point cloud in that Doppler interval is calculated based on the amplitude values ​​of all point clouds within that Doppler interval.

[0016] The energy percentage of a Doppler interval is calculated based on the ratio of the point cloud energy value of that Doppler interval to the sum of the point cloud energy values ​​of all Doppler intervals. The micro-Doppler array includes the energy percentage of each Doppler interval.

[0017] In one possible implementation, calculating the Doppler interval to which any point cloud in the associated point cloud belongs based on the radial velocity of that point cloud includes:

[0018] The index number of the Doppler interval to which the point cloud belongs is calculated according to the first preset formula, where the first preset formula is:

[0019]

[0020] Wherein, ind(i) represents the index number of the Doppler interval to which the i-th point cloud belongs, v(i) represents the radial velocity of the i-th point cloud. When the radial velocity is a positive velocity, the value of v(i) is positive, and when the radial velocity is a negative velocity, the value of v(i) is negative. k is a preset constant and is determined by the velocity resolution of the millimeter-wave radar. dopplersize represents the number of Doppler intervals in the micro-Doppler array. If the calculated value of ind(i) is not an integer and v(i) > 0, then the calculated value of ind(i) is rounded up to the smallest positive integer. If the calculated value of ind(i) is not an integer and v(i) < 0, then the calculated value of ind(i) is rounded down to the smallest positive integer.

[0021] In one possible implementation, determining whether it is a shower scene based on a micro-Doppler array of a consecutive first preset number of frames includes:

[0022] For each frame of micro-Doppler array, calculate the ratio of the total energy of all Doppler intervals belonging to the positive velocity interval to the total energy of the micro-Doppler array of that frame, and obtain the positive velocity energy ratio of the micro-Doppler array of that frame.

[0023] If, in the micro-Doppler array of the first preset number of frames, the proportion of positive velocity energy in the micro-Doppler array exceeding the third preset number of frames is greater than the preset proportion, or the proportion of positive velocity energy in the micro-Doppler array exceeding the preset proportion of frames is greater than the preset proportion, then it is determined to be a shower scene.

[0024] In one possible implementation, in the shower scene, obtaining a series of micro-Doppler arrays for a second preset number of consecutive frames, and calculating the energy Doppler distribution similarity of the associated point clouds for the second preset number of frames based on the series of micro-Doppler arrays for the second preset number of frames, includes:

[0025] For each frame of micro-Doppler array, the energy distribution entropy of the micro-Doppler array is calculated based on the proportion of each Doppler interval in the micro-Doppler array of that frame.

[0026] The standard deviation of the energy distribution entropy is calculated based on the energy distribution entropy of the micro-Doppler array for the second consecutive preset number of frames. The standard deviation of the energy distribution entropy is used to represent the energy Doppler distribution similarity of the associated point clouds for the second preset number of frames.

[0027] In one possible implementation, calculating the energy distribution entropy of each frame of micro-Doppler array based on the proportion of each Doppler interval in the frame includes:

[0028] The energy distribution entropy of the micro-Doppler array in this frame is calculated according to the second preset formula, which is:

[0029]

[0030] Wherein, entropy(m) represents the energy distribution entropy of the m-th frame micro-Doppler array, dopplersize represents the number of Doppler intervals in the micro-Doppler array, and dopplerArr(j) represents the energy percentage of the j-th Doppler interval in the m-th frame micro-Doppler array.

[0031] In one possible implementation, determining the target state based on the energy Doppler distribution similarity of the associated point clouds at the second preset frame number includes:

[0032] If the standard deviation of the energy distribution entropy is less than or equal to a preset threshold, then the target is determined to be in an inactive state.

[0033] If the standard deviation of the energy distribution entropy is greater than the preset threshold, then the target is determined to be in an active state.

[0034] Secondly, embodiments of the present invention provide a Doppler-based bathroom fall detection device, which is applied to a radar, wherein the radar is installed at a height not lower than the installation height of the shower head, and includes: a first acquisition module, a second acquisition module, a first determination module, a calculation module, a second determination module and a third determination module;

[0035] The first acquisition module is used to acquire the trajectory of the target and acquire the associated point cloud associated with the trajectory of the target frame by frame;

[0036] The second acquisition module is used to acquire the micro-Doppler array of the associated point cloud for each frame. The micro-Doppler array includes the energy distribution of the associated point cloud in multiple continuous but non-overlapping Doppler intervals. The Doppler intervals include positive velocity intervals and negative velocity intervals.

[0037] The first determining module is used to determine whether it is a shower scene based on the micro-Doppler array of a first preset number of consecutive frames;

[0038] The calculation module is used to obtain a continuous second preset number of micro-Doppler arrays in a shower scene, and calculate the energy Doppler distribution similarity of the associated point cloud for the corresponding second preset number of frames based on the continuous second preset number of micro-Doppler arrays.

[0039] The second determining module is used to determine the target state based on the energy Doppler distribution similarity of the associated point cloud of the second preset frame number, wherein the target state includes the target being in an active state or the target being in an inactive state.

[0040] The third determining module is used to determine that the target has fallen if the target changes from an active state to an inactive state.

[0041] Thirdly, embodiments of the present invention provide a radar, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method as described in the first aspect or any possible implementation of the first aspect above.

[0042] In one possible implementation, the radar is a millimeter-wave radar.

[0043] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0044] This invention, through pre-setting multiple Doppler intervals, analyzes the energy distribution of the point cloud associated with the target trajectory in each frame within each Doppler interval to obtain a micro-Doppler array for each frame's associated point cloud. Based on the characteristic that positive velocity energy dominates in a shower scene, the invention determines whether it is a shower scene by analyzing the micro-Doppler arrays of a first preset number of consecutive frames. Based on the characteristic that the inter-frame similarity of the micro-Doppler arrays is low when the target is active and high when the target is inactive, the invention determines whether the target is active by analyzing the micro-Doppler arrays of a second preset number of consecutive frames. If the target changes from an active to an inactive state, it is determined that the target has fallen in a shower scene. The method provided by this invention analyzes the person and water in a shower scene as a whole, solving the problem of water's influence on the detection of falls in existing technologies and improving the accuracy of fall detection in shower scenes. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a flowchart illustrating the implementation of the Doppler-based bathroom fall detection method provided in this embodiment of the invention.

[0047] Figure 2 This is a schematic diagram of the structure of the Doppler-based bathroom fall detection device provided in an embodiment of the present invention;

[0048] Figure 3 This is a schematic diagram of the radar provided in an embodiment of the present invention. Detailed Implementation

[0049] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0050] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0051] See Figure 1 The document illustrates a flowchart of the implementation of a Doppler-based bathroom fall detection method provided in an embodiment of the present invention, detailed below:

[0052] In step 101, the target's trajectory is acquired, and the associated point cloud associated with the target's trajectory is acquired frame by frame.

[0053] In this embodiment of the invention, optionally, the target's trajectory can be obtained through target detection.

[0054] The target referred to in the embodiments of the present invention may refer to a single target or one of multiple targets.

[0055] In one optional implementation, due to privacy concerns in shower scenarios and the need for timely assistance or alerts in case of falls when multiple people are present, a single-person shower scenario can be considered. In this embodiment, the target is a single object, and the trajectory within the bathroom is a stable, single trajectory. Each frame's associated point cloud is the point cloud associated with that single trajectory.

[0056] In this embodiment of the invention, the target's trajectory and each frame of associated point cloud are acquired using radar, such as millimeter-wave radar. The method provided in this embodiment of the invention is applied to a shower scenario, where the millimeter-wave radar is installed at a height no lower than the shower head's installation height.

[0057] In step 102, for each frame of associated point cloud, the micro-Doppler array of the frame of associated point cloud is obtained. The micro-Doppler array includes the energy distribution of the frame of associated point cloud in multiple continuous but non-overlapping Doppler intervals. The Doppler intervals include positive velocity intervals and negative velocity intervals.

[0058] After acquiring the associated point cloud for each frame, the velocity information of all point clouds in that frame is extracted. The velocity acquired by the millimeter-wave radar is the radial velocity, including both magnitude and direction. In this embodiment of the invention, the direction away from the millimeter-wave radar is defined as the positive direction, and the direction closer to the millimeter-wave radar is defined as the negative direction.

[0059] In this embodiment of the invention, multiple Doppler intervals are predefined. Two adjacent Doppler intervals are continuous but do not overlap. For example, 128 Doppler intervals are set, of which 64 consecutive Doppler intervals are negative velocity intervals, one middle Doppler interval is a zero velocity interval, and the other 63 consecutive Doppler intervals are positive velocity intervals.

[0060] The Doppler interval to which each point cloud belongs is determined based on the radial velocity of each point cloud. The energy distribution of the associated point cloud in each Doppler interval is statistically analyzed to obtain the energy distribution of the associated point cloud in multiple consecutive but non-overlapping Doppler intervals, and the micro-Doppler array of the associated point cloud in the frame is obtained.

[0061] Optionally, the size of each Doppler interval can be determined by the velocity resolution of the millimeter-wave radar, and the interval length of each Doppler interval is the same.

[0062] For example, if the velocity resolution of a millimeter-wave radar is 0.1 m / s, then the length of each Doppler interval is 0.1 m / s. If 0 velocity corresponds to one Doppler interval, 0 < v ≤ 0.1 m / s is the positive velocity interval adjacent to the Doppler interval corresponding to 0 velocity, and -0.1 ≤ v < 0 is the negative velocity interval adjacent to the Doppler interval corresponding to 0 velocity.

[0063] In one optional implementation, for any point cloud in the associated point cloud of the frame, the Doppler interval to which the point cloud belongs is calculated based on the radial velocity of the point cloud; for each Doppler interval, the point cloud energy value of the Doppler interval is calculated based on the amplitude values ​​of all point clouds within the Doppler interval; the energy proportion of the Doppler interval is calculated based on the ratio of the point cloud energy value of the Doppler interval to the point cloud energy values ​​of all Doppler intervals, and the micro-Doppler array includes the energy proportion of each Doppler interval.

[0064] In one optional implementation, the index number of the Doppler interval to which the point cloud belongs is calculated according to a first preset formula, wherein the first preset formula is:

[0065]

[0066] Wherein, ind(i) represents the index number of the Doppler interval to which the i-th point cloud belongs, v(i) represents the radial velocity of the i-th point cloud. When the radial velocity is a positive velocity, the value of v(i) is positive, and when the radial velocity is a negative velocity, the value of v(i) is negative. k is a preset constant and is determined by the velocity resolution of the millimeter-wave radar. dopplersize represents the number of Doppler intervals in the micro-Doppler array. If the calculated value of ind(i) is not an integer and v(i) > 0, then the calculated value of ind(i) is rounded up to the smallest positive integer. If the calculated value of ind(i) is not an integer and v(i) < 0, then the calculated value of ind(i) is rounded down to the smallest positive integer.

[0067] For example, by setting 128 Doppler intervals, with each interval corresponding to an index number of 1, 2, 3...128, a micro-Doppler array is constructed. This array has a size of 128, or an array length of 128. Specifically, Doppler intervals with indices 1 to 64 represent the negative velocity range, index 65 represents the zero velocity range, and indexes 66 to 128 represent the positive velocity range.

[0068] Using the above example, assuming k = 0.1 m / s and dopplersize = 128, then the index number corresponding to the 0 m / s speed range is 65, the index number corresponding to 0.05 m / s is 66, the index number corresponding to 0.1 m / s is also 66, the index number corresponding to 0.15 m / s is 67, the index number corresponding to -0.05 m / s is 64, and the index number corresponding to -0.1 m / s is also 64.

[0069] The index number calculated using the first preset formula is guaranteed to be a positive integer. This is one implementation provided by the present invention. Other implementations based on the concept of the present invention, such as index numbers with negative signs representing negative velocity ranges, are all within the protection scope of the present invention.

[0070] In this embodiment of the invention, optionally, since the square of the point cloud amplitude is linearly related to the point cloud energy, the point cloud amplitude can be used to characterize the point cloud energy. Other characteristic parameters of the point cloud can also be used to characterize the point cloud energy; this embodiment of the invention does not limit this approach.

[0071] Optionally, the point cloud amplitude is represented by amp. For any point cloud i, the energy value of the point cloud can be obtained by k(point(i).amp). 2 The expression represents the amplitude of point cloud i, where k is a preset coefficient and point(i).amp is the amplitude of point cloud i. In this embodiment of the invention, point cloud data is acquired through millimeter-wave radar, and the point cloud amplitude is data that can be directly obtained during the generation of point cloud data by the millimeter-wave radar.

[0072] For each Doppler interval, the sum of the energy of all point clouds within that interval is calculated to obtain the point cloud energy value for that interval. The point cloud energy value for each Doppler interval is then normalized by calculating the ratio of the point cloud energy value of each interval to the sum of the energy values ​​of all Doppler intervals, thus obtaining the energy proportion of each Doppler interval. In a micro-Doppler array of a frame of associated point clouds, the data stored in that micro-Doppler array represents the energy proportion of each Doppler interval.

[0073] In step 103, the micro-Doppler array of a first preset number of consecutive frames is used to determine whether it is a shower scene.

[0074] In this embodiment of the invention, a micro-Doppler spectrum or micro-Doppler image can be constructed by splicing together a series of micro-Doppler arrays for a first preset number of frames in time. The micro-Doppler spectrum or micro-Doppler image can be analyzed to determine whether it is a shower scene. Alternatively, the data in the micro-Doppler array for the first preset number of frames can be directly analyzed to determine whether it is a shower scene. This embodiment of the invention does not limit this approach.

[0075] In this embodiment of the invention, the installation height of the millimeter-wave radar is not lower than the installation height of the shower head. Preferably, the installation height of the millimeter-wave radar is the same as the installation height of the shower head, and the area above the shower head is not detected. Alternatively, if the installation height of the millimeter-wave radar is higher than the installation height of the shower head, the detected point cloud data above the shower head is deleted according to the relative position of the millimeter-wave radar and the shower head.

[0076] Because shower scenes involve numerous point clouds of water, and the person showering is integrated with the water, making point cloud data difficult to distinguish, existing fall detection methods based on millimeter-wave radar have low accuracy. To address this issue, in this embodiment of the invention, the person and water in a shower scene are treated as a whole. Considering the large number of point clouds in the water and the downward movement due to gravity, the shower scene is represented in the micro-Doppler array as follows: point cloud energy in the direction away from the millimeter-wave radar, i.e., the positive velocity direction in this embodiment, dominates. For a continuous first preset number of micro-Doppler array frames, if the ratio of the micro-Doppler array dominated by the positive velocity point cloud energy or the number of frames reaches a preset threshold, it can be determined that the person is in a shower scene.

[0077] In one optional implementation, for each frame of micro-Doppler array, the ratio of the sum of the energy of all Doppler intervals belonging to the positive velocity interval to the total energy of the micro-Doppler array of that frame is calculated to obtain the positive velocity energy ratio of the micro-Doppler array of that frame; if, in the micro-Doppler array of the first preset number of frames, the positive velocity energy ratio of the micro-Doppler array of the third preset number of frames is greater than the preset ratio, or the positive velocity energy ratio of the micro-Doppler array of the first preset number of frames is greater than the preset ratio, then it is determined to be a shower scene.

[0078] Optionally, if the index number of the Doppler interval to which each point cloud belongs is calculated based on the first preset formula, then the Doppler interval with index number 0.5×dopplersize+1 to index number dopplersize corresponds to the positive velocity interval. The ratio of the sum of the point cloud energy of the Doppler interval with index number 0.5×dopplersize+1 to index number dopplersize to the total energy of the micro-Doppler array of that frame is calculated to obtain the positive velocity energy ratio of the micro-Doppler array of that frame.

[0079] If the point cloud energy value of each Doppler interval is normalized in step 102, then the data stored in the micro-Doppler array of this frame is the energy proportion of each Doppler interval. After determining that the Doppler intervals with index number 0.5×dopplersize+1 to index number dopplersize correspond to the positive velocity interval, the positive velocity energy proportion of the micro-Doppler array of this frame can be obtained by adding the values ​​of the Doppler intervals with index number 0.5×dopplersize+1 to index number dopplersize in the micro-Doppler array.

[0080] Optionally, the positive velocity energy percentage of the micro-Doppler array is represented by a ratio. If the ratio of a frame is greater than a preset ratio (e.g., 0.8), the micro-Doppler data flag for that frame is 1; otherwise, it is 0. A preset buffer space is used to store the flags for each frame. Assuming the buffer size is BUFSIZE, for example, if the first preset frame count is 500 frames, then BUFSIZE = 500. When the buffer space is full, a determination is made as to whether it is a shower scene. Assuming the preset ratio is 95%, if more than 95% of the flag data in the buffer space is 1, it is determined to be a shower scene. Alternatively, assuming the third preset frame count is 450 frames, if the number of flag data with 1 in the buffer space exceeds 450, it is determined to be a shower scene.

[0081] Correspondingly, the energy proportion in the negative velocity range can also be analyzed. In the shower scene, the point cloud energy in the negative velocity direction is not dominant, so the energy proportion of the negative velocity direction is low. For a single frame micro-Doppler array, the negative velocity energy proportion of the micro-Doppler array in that frame is obtained. If, in the micro-Doppler array of the first preset number of frames, the negative velocity energy proportion of the micro-Doppler array of more than one preset number of frames is less than a preset proportion, or the negative velocity energy proportion of the micro-Doppler array of more than one preset proportion of frames is less than a preset proportion, then it is determined to be a shower scene.

[0082] In step 104, under the shower scene, the micro-Doppler array of the second preset number of consecutive frames is obtained, and the energy Doppler distribution similarity of the associated point cloud of the corresponding second preset number of frames is calculated based on the micro-Doppler array of the second preset number of consecutive frames.

[0083] In a shower scene, when a person falls, the signal is generally weak. At this time, the water point cloud dominates, and due to the absence of human activity, the data distribution of the micro-Doppler arrays is similar across frames. When the target is active, the data distribution of the micro-Doppler arrays varies significantly across frames due to the randomness of human movement. Therefore, in this step, the similarity of the micro-Doppler array data distribution across a second preset number of frames is used as the basis for determining the target's state in the shower scene.

[0084] In this embodiment of the invention, the point cloud energy value in each Doppler interval is normalized using the method provided in step 102 to obtain the energy ratio of each Doppler interval. That is, the data stored in the micro-Doppler array in this step is the energy ratio of each Doppler interval.

[0085] In this embodiment of the invention, the similarity between every two micro-Doppler arrays in the second preset number of micro-Doppler arrays can be calculated using existing methods for calculating array similarity, resulting in a similarity sequence. The standard deviation or variance of the similarity sequence is then calculated to obtain the energy Doppler distribution similarity of the associated point clouds in the second preset number of frames. The smaller the standard deviation or variance of the similarity sequence, the higher the energy Doppler distribution similarity of the associated point clouds in the second preset number of frames. Optionally, when the standard deviation or variance is less than a preset threshold, the inter-frame micro-Doppler array similarity is high, which meets the discrimination feature that the target is in an inactive state. In this case, the target state is determined to be inactive; otherwise, it is active.

[0086] Entropy is a measure of uncertainty. In one alternative implementation, since the data stored in the micro-Doppler array is the energy percentage of each Doppler interval, and the energy percentage of each Doppler interval can represent the probability of energy occurrence, the data stored in the micro-Doppler array can be used as the raw value for entropy calculation. Based on this, this embodiment of the invention uses entropy values ​​to analyze the dispersion of energy distribution in the micro-Doppler array.

[0087] For each frame of micro-Doppler array, the energy distribution entropy of the micro-Doppler array is calculated based on the proportion of each Doppler interval in the micro-Doppler array of that frame; the standard deviation of the energy distribution entropy is calculated based on the energy distribution entropy of the micro-Doppler arrays of the second preset number of consecutive frames. The standard deviation of the energy distribution entropy is used to represent the energy Doppler distribution similarity of the associated point clouds of the second preset number of frames.

[0088] Optionally, the energy distribution entropy of the micro-Doppler array in this frame is calculated according to a second preset formula, which is:

[0089]

[0090] Where entropy(m) represents the energy distribution entropy of the m-th frame micro-Doppler array, dOpplersize represents the number of Doppler intervals in the micro-Doppler array, and dOpplerArr(j) represents the energy percentage of the j-th Doppler interval in the m-th frame micro-Doppler array.

[0091] In step 105, the target state is determined based on the energy Doppler distribution similarity of the associated point cloud at the second preset frame number. The target state includes whether the target is in an active state or the target is in an inactive state.

[0092] In this embodiment of the invention, since the data distribution of unmanned features and weak human life signals is similar, the target in the embodiment of the invention is in an inactive state, including an unmanned state and a state with weak human life signals.

[0093] Optionally, a threshold value for the energy Doppler distribution similarity can be set. If the energy Doppler distribution similarity of the associated point cloud in the second preset frame is higher than the threshold value, the target is in an inactive state; otherwise, the target is in an active state.

[0094] In one optional implementation, the energy distribution entropy of the micro-Doppler arrays over time is accumulated to form an entropy curve, and the standard deviation is extracted as a measure of jitter. When the target is active, especially during vigorous activity such as falling, the ratio of positive to negative energy in the micro-Doppler arrays is stable due to the randomness of the target's movement, resulting in low similarity between micro-Doppler array frames and high curve jitter. When the target is inactive, i.e., when there is no one or the vital signs are weak, the similarity between micro-Doppler array frames is high, and the curve jitter is low. The target state is determined by setting a threshold for the standard deviation.

[0095] Optionally, if the standard deviation of the energy distribution entropy is less than or equal to a preset threshold, the target is determined to be in an inactive state; if the standard deviation of the energy distribution entropy is greater than the preset threshold, the target is determined to be in an active state.

[0096] Furthermore, in shower scenarios where no one is present or a person is in danger (i.e., the target is inactive), the similarity between micro-Doppler arrays is high, and the micro-Doppler arrays exhibit a consistently dominant and stable positive velocity energy. To improve discrimination accuracy, an additional discrimination condition is added when determining the target state: in the micro-Doppler arrays of the second preset frame number, the proportion of positive velocity energy in the micro-Doppler arrays for more than four consecutive preset frame numbers is greater than a preset proportion. The fourth preset frame number is less than or equal to the second preset frame number.

[0097] Alternatively, to improve the discrimination accuracy, the preset threshold corresponding to the standard deviation of the energy distribution entropy is divided into a first preset threshold and a second preset threshold. The first preset threshold is less than the second preset threshold. If the standard deviation of the energy distribution entropy is less than or equal to the first preset threshold, the target state is determined to be an inactive state. If the standard deviation of the energy distribution entropy is greater than the second preset threshold, the target state is determined to be an active state. If the standard deviation of the energy distribution entropy is greater than the first preset threshold and less than or equal to the second preset threshold, the target state can be marked as other states.

[0098] In step 106, if the target changes from an active state to an inactive state, it is determined that the target has fallen.

[0099] In this embodiment of the invention, a judgment is performed every second preset number of frames, and a target state is obtained each time. If the target state changes from an active state to an inactive state, it is determined that the target has fallen, and an alarm is triggered in a timely manner.

[0100] In this embodiment of the invention, if, as described in step 105, the target state may include other states besides the target being in an active state and the target being in an inactive state, then the target changing from an active state to an inactive state may be two consecutive judgment results, or after obtaining the judgment result that the target is in an active state, other states may be spaced out, and then the judgment result that the target is in an inactive state may be obtained, which also satisfies the fall alarm condition.

[0101] The method provided in the embodiments of the present invention has at least the following advantages:

[0102] First, it does not need to be worn on the body; it can be installed in the corner of the bathroom to provide real-time safety protection for people in the bathroom.

[0103] Second, in this embodiment of the invention, point cloud data is acquired by millimeter-wave radar. The millimeter-wave radar is located in the overlapping region of far-infrared and microwave, so there will be no privacy leakage problem.

[0104] Third, the data processing stage utilizes micro-Doppler information feature extraction, which significantly reduces water interference and is computationally convenient, highly portable, and cost-effective.

[0105] Fourth, compared to conventional millimeter-wave solutions, it has higher alarm accuracy and stronger reliability in shower scenarios.

[0106] This invention, through pre-setting multiple Doppler intervals, analyzes the energy distribution of the point cloud associated with the target trajectory in each frame within each Doppler interval to obtain a micro-Doppler array for each frame's associated point cloud. Based on the characteristic that positive velocity energy dominates in a shower scene, the invention determines whether it is a shower scene by analyzing the micro-Doppler arrays of a first preset number of consecutive frames. Based on the characteristic that the inter-frame similarity of the micro-Doppler arrays is low when the target is active and high when the target is inactive, the invention determines whether the target is active by analyzing the micro-Doppler arrays of a second preset number of consecutive frames. If the target changes from an active to an inactive state, it is determined that the target has fallen in a shower scene. The method provided by this invention analyzes the person and water in a shower scene as a whole, solving the problem of water's influence on the detection of falls in existing technologies and improving the accuracy of fall detection in shower scenes.

[0107] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0108] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0109] Figure 2 A schematic diagram of a Doppler-based bathroom fall detection device provided in an embodiment of the present invention is shown. For ease of explanation, only the parts relevant to the embodiment of the present invention are shown, and are described in detail below:

[0110] like Figure 2 As shown, the Doppler-based bathroom fall detection device 2 includes: a first acquisition module 21, a second acquisition module 22, a first determination module 23, a calculation module 24, a second determination module 25, and a third determination module 26;

[0111] The first acquisition module 21 is used to acquire the target's trajectory and acquire the associated point cloud associated with the target's trajectory frame by frame;

[0112] The second acquisition module 22 is used to acquire the micro-Doppler array of the associated point cloud for each frame. The micro-Doppler array includes the energy distribution of the associated point cloud in multiple continuous but non-overlapping Doppler intervals. The Doppler intervals include positive velocity intervals and negative velocity intervals.

[0113] The first determining module 23 is used to determine whether it is a shower scene based on the micro-Doppler array of a continuous first preset number of frames;

[0114] The calculation module 24 is used to obtain the micro-Doppler array of a continuous second preset number of frames in the shower scene, and calculate the energy Doppler distribution similarity of the associated point cloud of the corresponding second preset number of frames based on the micro-Doppler array of the continuous second preset number of frames.

[0115] The second determining module 25 is used to determine the target state based on the energy Doppler distribution similarity of the associated point cloud of the second preset frame number. The target state includes whether the target is in an active state or the target is in an inactive state.

[0116] The third determining module 26 is used to determine if the target has fallen if it changes from an active state to an inactive state.

[0117] This invention, through pre-setting multiple Doppler intervals, analyzes the energy distribution of the point cloud associated with the target trajectory in each frame within each Doppler interval to obtain a micro-Doppler array for each frame's associated point cloud. Based on the characteristic that positive velocity energy dominates in a shower scene, the invention determines whether it is a shower scene by analyzing the micro-Doppler arrays of a first preset number of consecutive frames. Based on the characteristic that the inter-frame similarity of the micro-Doppler arrays is low when the target is active and high when the target is inactive, the invention determines whether the target is active by analyzing the micro-Doppler arrays of a second preset number of consecutive frames. If the target changes from an active to an inactive state, it is determined that the target has fallen in a shower scene. The method provided by this invention analyzes the person and water in a shower scene as a whole, solving the problem of water's influence on the detection of falls in existing technologies and improving the accuracy of fall detection in shower scenes.

[0118] In one possible implementation, the second acquisition module 22 is used for:

[0119] For any point cloud in the associated point cloud of this frame, calculate the Doppler interval to which the point cloud belongs based on the radial velocity of the point cloud;

[0120] For each Doppler interval, the energy value of the point cloud in that Doppler interval is calculated based on the amplitude values ​​of all point clouds within that Doppler interval.

[0121] The energy percentage of a Doppler interval is calculated based on the ratio of the point cloud energy value of that interval to the sum of the point cloud energy values ​​of all Doppler intervals. The micro-Doppler array includes the energy percentage of each Doppler interval.

[0122] In one possible implementation, the second acquisition module 22 is used for:

[0123] The index number of the Doppler interval to which the point cloud belongs is calculated according to the first preset formula, which is:

[0124]

[0125] Wherein, ind(i) represents the index number of the Doppler interval to which the i-th point cloud belongs, v(i) represents the radial velocity of the i-th point cloud. When the radial velocity is in the positive direction, the value of v(i) is positive, and when the radial velocity is in the negative direction, the value of v(i) is negative. k is a preset constant and is determined by the velocity resolution of the millimeter-wave radar. dopplersize represents the number of Doppler intervals in the micro-Doppler array. If the calculated value of ind(i) is not an integer and v(i) > 0, then the calculated value of ind(i) is rounded up to the smallest positive integer. If the calculated value of ind(i) is not an integer and v(i) < 0, then the calculated value of ind(i) is rounded down to the smallest positive integer.

[0126] In one possible implementation, the first determining module 23 is used for:

[0127] For each frame of micro-Doppler array, calculate the ratio of the total energy of all Doppler intervals belonging to the positive velocity interval to the total energy of the micro-Doppler array of that frame, and obtain the positive velocity energy ratio of the micro-Doppler array of that frame.

[0128] If, in the micro-Doppler array of the first preset number of frames, the proportion of positive velocity energy in the micro-Doppler array of the third preset number of frames is greater than the preset proportion, or the proportion of positive velocity energy in the micro-Doppler array of the preset proportion of frames is greater than the preset proportion, then it is determined to be a shower scene.

[0129] In one possible implementation, the computing module 24 is used for:

[0130] For each frame of micro-Doppler array, the energy distribution entropy of the micro-Doppler array is calculated based on the proportion of each Doppler interval in the micro-Doppler array of that frame.

[0131] The standard deviation of the energy distribution entropy is calculated based on the energy distribution entropy of the micro-Doppler array for a consecutive second preset number of frames. The standard deviation of the energy distribution entropy is used to represent the energy Doppler distribution similarity of the associated point clouds for the second preset number of frames.

[0132] In one possible implementation, the computing module 24 is used for:

[0133] The energy distribution entropy of the micro-Doppler array in this frame is calculated according to the second preset formula, which is as follows:

[0134]

[0135] Where entropy(m) represents the energy distribution entropy of the m-th frame micro-Doppler array, dopplersize represents the number of Doppler intervals in the micro-Doppler array, and dopplerArr(j) represents the energy percentage of the j-th Doppler interval in the m-th frame micro-Doppler array.

[0136] In one possible implementation, the second determining module 25 is used for:

[0137] If the standard deviation of the energy distribution entropy is less than or equal to a preset threshold, the target is determined to be in an inactive state.

[0138] If the standard deviation of the energy distribution entropy is greater than a preset threshold, the target is determined to be active.

[0139] The Doppler-based bathroom fall detection device provided in this embodiment can be used to execute the above-described Doppler-based bathroom fall detection method embodiment. Its implementation principle and technical effect are similar, and will not be described again here.

[0140] Figure 3 This is a schematic diagram of a radar provided according to an embodiment of the present invention. Figure 3 As shown, the radar 3 in this embodiment includes: a processor 30, a memory 31, and a computer program 32 stored in the memory 31 and executable on the processor 30. When the processor 30 executes the computer program 32, it implements the steps described in the various embodiments of the Doppler-based bathroom fall detection method, for example... Figure 2 Steps 101 to 106 are shown. Alternatively, when the processor 30 executes the computer program 32, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 2 The functions of modules 21 to 26 are shown.

[0141] For example, the computer program 32 can be divided into one or more modules / units, which are stored in the memory 31 and executed by the processor 30 to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program 32 in the radar 3.

[0142] The radar 3 may be a millimeter-wave radar. The radar 3 may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will understand that... Figure 3 This is merely an example of radar 3 and does not constitute a limitation on radar 3. It may include more or fewer components than shown, or combine certain components, or different components. For example, the radar may also include input / output devices, network access devices, buses, etc.

[0143] The processor 30 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0144] The memory 31 can be an internal storage unit of the radar 3, such as a hard disk or memory of the radar 3. The memory 31 can also be an external storage device of the radar 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the radar 3. Furthermore, the memory 31 can include both internal storage units and external storage devices of the radar 3. The memory 31 is used to store the computer program and other programs and data required by the radar. The memory 31 can also be used to temporarily store data that has been output or will be output.

[0145] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0146] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0147] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0148] In the embodiments provided by this invention, it should be understood that the disclosed apparatus / radar and method can be implemented in other ways. For example, the apparatus / radar embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0149] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0150] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0151] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various embodiments of the Doppler-based bathroom fall detection method described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0152] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A Doppler-based method for detecting falls in bathrooms, characterized in that, Applied to a radar, wherein the installation height of the radar is not lower than the installation height of the shower head, the detection method includes: Acquire the target's trajectory and acquire the associated point cloud related to the target's trajectory frame by frame; For each frame of associated point cloud, the micro-Doppler array of the associated point cloud of that frame is obtained, including: for any point cloud in the associated point cloud of that frame, calculating the Doppler interval to which the point cloud belongs based on the radial velocity of the point cloud; for each Doppler interval, calculating the point cloud energy value of the Doppler interval based on the amplitude values ​​of all point clouds within the Doppler interval; calculating the energy proportion of the Doppler interval based on the ratio of the point cloud energy value of the Doppler interval to the sum of the point cloud energy values ​​of all Doppler intervals, wherein the micro-Doppler array includes the energy proportion of each Doppler interval; the micro-Doppler array includes the energy distribution of the associated point cloud of that frame in multiple consecutive but non-overlapping Doppler intervals, wherein the Doppler interval includes positive velocity intervals and negative velocity intervals; Based on the micro-Doppler array of a consecutive first preset number of frames, determine whether it is a shower scene; In a shower scene, a series of consecutive micro-Doppler arrays for a second preset number of frames are acquired, and the energy Doppler distribution similarity of the associated point clouds for the corresponding second preset number of frames is calculated based on the series of consecutive micro-Doppler arrays for the second preset number of frames. This includes: for each frame of micro-Doppler array, calculating the energy distribution entropy of the frame's micro-Doppler array based on the proportion of each Doppler interval in the frame's micro-Doppler array; and calculating the standard deviation of the energy distribution entropy based on the energy distribution entropy of the series of consecutive micro-Doppler arrays for the second preset number of frames, wherein the standard deviation of the energy distribution entropy is used to represent the energy Doppler distribution similarity of the associated point clouds for the second preset number of frames. The target state is determined based on the energy Doppler distribution similarity of the associated point cloud of the second preset frame number. The target state includes whether the target is in an active state or the target is in an inactive state. If the target changes from an active state to an inactive state, then the target is determined to have fallen.

2. The method according to claim 1, characterized in that, The step of calculating the Doppler interval to which any point cloud in the associated point cloud belongs based on the radial velocity of the point cloud includes: The index number of the Doppler interval to which the point cloud belongs is calculated according to the first preset formula, where the first preset formula is: in, Used to indicate the first The index number of the Doppler interval to which the point cloud belongs. Used to indicate the first The radial velocity of a point cloud, when the radial velocity is in the positive direction. The value is positive when the radial velocity is a negative velocity. The value is negative. This is a preset constant and is determined by the velocity resolution of the millimeter-wave radar. Used to represent the number of Doppler intervals in the micro-Doppler array, if The calculated value is not an integer, and >0, then for The calculated value is rounded up to the smallest positive integer. The calculated value is not an integer, and <0, then for The calculated value is rounded down to the smallest positive integer.

3. The method according to claim 1, characterized in that, The step of determining whether it is a shower scene based on the micro-Doppler array of a consecutive first preset number of frames includes: For each frame of micro-Doppler array, calculate the ratio of the total energy of all Doppler intervals belonging to the positive velocity interval to the total energy of the micro-Doppler array of that frame, and obtain the positive velocity energy ratio of the micro-Doppler array of that frame. If, in the micro-Doppler array of the first preset number of frames, the proportion of positive velocity energy in the micro-Doppler array exceeding the third preset number of frames is greater than the preset proportion, or the proportion of positive velocity energy in the micro-Doppler array exceeding the preset proportion of frames is greater than the preset proportion, then it is determined to be a shower scene.

4. The method according to claim 1, characterized in that, For each frame of the micro-Doppler array, the energy distribution entropy of the micro-Doppler array is calculated based on the proportion of each Doppler interval in the frame, including: The energy distribution entropy of the micro-Doppler array in this frame is calculated according to the second preset formula, which is: in, Used to indicate the first The energy distribution entropy of the frame micro-Doppler array Used to indicate the number of Doppler intervals in the microDoppler array. Used to indicate the first The first frame of the micro Doppler array The energy percentage of each Doppler interval.

5. The method according to claim 1, characterized in that, The step of determining the target state based on the energy Doppler distribution similarity of the associated point cloud according to the second preset frame number includes: If the standard deviation of the energy distribution entropy is less than or equal to a preset threshold, then the target is determined to be in an inactive state. If the standard deviation of the energy distribution entropy is greater than the preset threshold, then the target is determined to be in an active state.

6. A Doppler-based bathroom fall detection device, characterized in that, The radar where the detection device is located is installed at a height no lower than the installation height of the shower head. The detection device includes: a first acquisition module, a second acquisition module, a first determination module, a calculation module, a second determination module, and a third determination module. The first acquisition module is used to acquire the trajectory of the target and acquire the associated point cloud associated with the trajectory of the target frame by frame; The second acquisition module is used to acquire a micro-Doppler array of the associated point cloud for each frame, including: for any point cloud in the associated point cloud of the frame, calculating the Doppler interval to which the point cloud belongs based on the radial velocity of the point cloud; for each Doppler interval, calculating the point cloud energy value of the Doppler interval based on the amplitude values ​​of all point clouds within the Doppler interval; calculating the energy proportion of the Doppler interval based on the ratio of the point cloud energy value of the Doppler interval to the sum of the point cloud energy values ​​of all Doppler intervals, wherein the micro-Doppler array includes the energy proportion of each Doppler interval; the micro-Doppler array includes the energy distribution of the associated point cloud of the frame in multiple consecutive but non-overlapping Doppler intervals, wherein the Doppler interval includes a positive velocity interval and a negative velocity interval; The first determining module is used to determine whether it is a shower scene based on the micro-Doppler array of a first preset number of consecutive frames; The calculation module is used to acquire a continuous second preset number of micro-Doppler arrays in a shower scene, and calculate the energy Doppler distribution similarity of the associated point clouds for the corresponding second preset number of frames based on the continuous second preset number of micro-Doppler arrays; including: for each frame of micro-Doppler array, calculating the energy distribution entropy of the frame of micro-Doppler array based on the proportion of each Doppler interval in the frame of micro-Doppler array; and calculating the standard deviation of the energy distribution entropy based on the energy distribution entropy of the continuous second preset number of micro-Doppler arrays, wherein the standard deviation of the energy distribution entropy is used to represent the energy Doppler distribution similarity of the associated point clouds for the second preset number of frames; The second determining module is used to determine the target state based on the energy Doppler distribution similarity of the associated point cloud of the second preset frame number, wherein the target state includes the target being in an active state or the target being in an inactive state. The third determining module is used to determine that the target has fallen if the target changes from an active state to an inactive state.

7. A radar comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 5 above.

8. The radar according to claim 7, characterized in that, The radar in question is a millimeter-wave radar.

Images