Fall detection method and device

Abstract

To provide a fall detection method and device.SOLUTION: The fall detection method includes the steps of: performing fall detection on the basis of radar point group information in a first stage, and shifting to a second stage when no fall event is detected; performing fall detection on the basis of the radar point group information and radar angle FFT information in the second stage, and shifting to a third stage when no fall event is detected; and performing fall detection on the basis of the radar angle FFT information in the third stage. The fall detection is performed using the radar point group information and the radar angle FFT information at different stages where the fall may occur to improve the speed and reliability of the fall detection by determining whether or not the fall event occurs.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to the field of vital signs detection, and in particular to a method and apparatus for detecting falls. [Background technology]

[0002] Currently, the number and proportion of the elderly population is increasing, and the global population is entering an aging phase. Paying attention to the physical health and quality of life of the elderly and resolving the problems they face are of great social and economic significance. Falls cause significant damage to the physical health and life safety of the elderly. According to statistics from the United Nations Health Organization, falls are the second leading cause of death from unintentional injuries worldwide. Fall detection technology can help quickly detect falls, facilitate rapid rescue, and prevent injuries from worsening.

[0003] Common solutions for fall detection include video-based detection technology and wearable device-based detection methods. Video-based fall detection technology uses intelligent detection methods to analyze video data to detect fall events. However, because good lighting conditions are required to acquire video data, this technology cannot be used in dark environments at night. Furthermore, cameras significantly invade privacy and cannot be placed in private environments such as bedrooms or bathrooms. Wearable device-based detection methods use sensors such as gyroscopes and accelerometers to analyze the user's movement characteristics to determine whether a fall has occurred. However, this method requires the user to wear the device to function, and user acceptance is low due to issues with the comfort of wearable devices and frequent charging.

[0004] The above description of the technical background is intended to provide a clear and complete understanding of the aspects of the present invention, and is described for the understanding of persons skilled in the art. These technical solutions are merely described as part of the background technology of the present invention, and are not well known by persons skilled in the art. Summary of the Invention [Problem to be solved by the invention]

[0005] According to the inventors' findings, millimeter-wave radar-based fall detection technology has the advantages of not requiring users to wear a device, not requiring lighting, and protecting privacy. It can be applied to private places such as bedrooms and bathrooms, and has good market potential. Radar-based fall detection technology can analyze the movement characteristics of the human body and determine whether a fall has occurred using methods such as machine learning or template matching. It can also determine whether a fall has occurred by detecting whether a person is lying on the ground based on radar signal fluctuations. However, the former method is prone to missed detections, while the latter method takes a long time to detect and is prone to false positives. Furthermore, radar can provide limited information, making it difficult to quickly and reliably detect falls.

[0006] In view of at least one of the above problems or similar problems, embodiments of the present invention provide a fall detection method and device that can improve the speed and reliability of fall detection by performing fall detection using radar point cloud information and radar angle FFT information at different stages when a fall may occur and determining whether a fall event has occurred. [Means for solving the problem]

[0007] In one aspect of an embodiment of the present invention, there is provided a fall detection device including a detection unit that, in a first stage, performs fall detection based on radar point cloud information, and if a fall event is not detected, proceeds to a second stage, performs fall detection based on the radar point cloud information and radar angle FFT information in the second stage, and if a fall event is not detected, proceeds to a third stage, and in the third stage, performs fall detection based on the radar angle FFT information.

[0008] Another aspect of an embodiment of the present invention provides a fall detection method including: in a first stage, performing fall detection based on radar point cloud information, and transitioning to a second stage if a fall event is not detected; in the second stage, performing fall detection based on the radar point cloud information and radar angle FFT information, and transitioning to a third stage if a fall event is not detected; and in the third stage, performing fall detection based on the radar angle FFT information.

[0009] Another aspect of an embodiment of the present invention provides a computing device including a memory having a computer program stored therein and a processor, the processor configured to execute the computer program to implement the above method.

[0010] Another aspect of an embodiment of the present invention provides a storage medium having a computer-readable program stored thereon, the computer-readable program causing a computer to execute the above-described method.

[0011] One of the advantageous effects of the embodiment of the present invention is as follows: According to the embodiment of the present invention, the speed and reliability of fall detection can be improved by performing fall detection using radar point cloud information and radar angle FFT information at different stages when a fall may occur, and determining whether a fall event has occurred.

[0012] Certain embodiments of the present invention are disclosed in detail, as set forth in the following description and in the drawings, and are indicative of ways in which the principles of the present invention may be employed. However, the embodiments of the present invention are not intended to be limiting in scope. The present invention encompasses various alterations, modifications, and equivalents within the spirit and content of the appended claims.

[0013] Features described and / or shown in one embodiment may be used in the same or similar manner in one or more other embodiments, may be combined with features in the other embodiments, or may substitute for features in the other embodiments.

[0014] It should be noted that the term "comprise / have" when used in this context means the presence of a feature, element, step or component, and does not exclude the presence or addition of one or more other features, elements, steps or components. [Brief explanation of the drawings]

[0015] Elements and features illustrated in one drawing or embodiment of an example of the invention may be combined with elements and features shown in one or more other drawings or embodiments. In the drawings, like reference numerals represent corresponding components in multiple drawings and may represent corresponding components used in multiple aspects.

[0016] The drawings included herein are for understanding the embodiments of the present invention, constitute a part of this specification, and are for illustrating the embodiments of the present invention, and together with the written description, explain the principles of the present invention. Note that the drawings described herein are merely for illustrating the embodiments of the present invention, and those skilled in the art can easily derive other drawings based on these drawings. [Figure 1] 1 is a schematic diagram of an example of a fall detection method according to an embodiment of the present invention; [Figure 2] 1 is a schematic diagram of the stages of fall detection according to a method according to an embodiment of the invention; [Figure 3] FIG. 1 is a schematic diagram of an example of radar point cloud based fall detection in a first stage. [Figure 4] FIG. 10 is a schematic diagram of an example of fall detection based on radar point cloud information and radar angle FFT information in the second stage. [Figure 5] FIG. 10 is a schematic diagram of an example of fall detection based on radar angle FFT information in the third stage. [Figure 6]2 is a schematic diagram of an example of a fall detection process of a method according to an embodiment of the present invention; [Figure 7] 1 is a schematic diagram of an example of a fall detection device according to an embodiment of the present invention. [Figure 8] FIG. 1 is a schematic diagram of an example of a computing device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0017] These and other features of the present invention will become apparent from the drawings and the following description. The specification and drawings disclose specific embodiments of the present invention, i.e., some embodiments consistent with the principles of the present invention. It should be noted that the present invention is not limited to the described embodiments, but includes all modifications, variations, and equivalents within the scope of the appended claims.

[0018] In the embodiments of the present invention, the terms "first" and "second" are used to distinguish different elements by name and do not refer to the spatial arrangement or temporal order of these elements, and these elements are not limited to these terms. The term "and / or" includes any one or more of the listed terms and combinations thereof. The terms "including," "including," and "having" refer to the presence of stated features, elements, elements, or components, but do not exclude the presence or addition of one or more other features, elements, elements, or components.

[0019] In the embodiments of the present invention, the singular forms "a," "the," etc. include the plural and mean "a kind" or "a class," and are not limited to "one." Furthermore, the term "said" includes both the singular and the plural unless the context clearly dictates otherwise. Furthermore, the term "according to" means "at least partially depending on," and the term "based on" means "at least partially based on," unless the context clearly dictates otherwise.

[0020] Hereinafter, each embodiment of the present invention will be described with reference to the drawings.

[0021] Example 1 An embodiment of the present invention provides a fall detection method, more specifically, a multi-stage fall detection method based on millimeter-wave radar, which uses millimeter-wave radar point cloud information and radar angle FFT (Fast Fourier Transform) information to perform detection at different stages of a fall event.

[0022] 1 is a schematic diagram of an example of a fall detection method according to an embodiment of the present invention. As shown in FIG. 1, the method includes the following steps:

[0023] Step 110: In the first stage, a fall is detected based on the radar point cloud information, and if a fall event is not detected, the process proceeds to the second stage.

[0024] Step 120: In the second stage, a fall is detected based on the radar point cloud information and the radar angle FFT information, and if a fall event is not detected, the process proceeds to the third stage.

[0025] Step 130: In the third stage, fall detection is performed based on the radar angle FFT information.

[0026] According to an embodiment of the present invention, the speed and reliability of fall detection can be improved by performing fall detection using radar point cloud information and radar angle FFT information at different stages when a fall may occur and determining whether a fall event has occurred.

[0027] In an embodiment of the present invention, a radar periodically emits a radar signal into space, receives a reflected signal, and processes the reflected signal to detect an object that has reflected the radar signal. The reflected signal received by the radar is a superposition of signals reflected from different objects, and by performing processing such as range FFT, Doppler FFT, and radar angle FFT, the reflected signal can be distinguished in multiple dimensions such as range, Doppler, and angle.

[0028] In the embodiment of the present invention, the radar angle FFT information means the result of performing the above processing on the reflected signal received by the radar, and S={sx,y,z}, where s x,y,z represents the radar signal emitted from the spatial location (x,y,z).

[0029] In an embodiment of the present invention, the point cloud of one frame of the radar is expressed as P={(x i ,y i ,z i ,v i ), 1≦i≦n}, where (x i ,y i ,z i ) are the spatial coordinates of the point cloud, and z i is the height of the point cloud relative to the ground, and v i is the Doppler velocity of the point cloud, and n is the number of point clouds. In the embodiment of the present invention, a point cloud reflecting a moving target is called an effective point cloud. Therefore, if a radar cannot generate an effective point cloud, the radar cannot detect a moving target within its coverage area. On the other hand, if a radar can generate an effective point cloud, the radar can detect a moving target within its coverage area.

[0030] In the embodiments of the present invention, a millimeter wave radar, for example, an FCMW (Frequency Modulated Continuous Wave) radar, is used as an example, but the present invention is not limited to this, and the radar may be any other type of radar as long as it is capable of generating point cloud information.

[0031] FIG. 2 is a schematic diagram of the stages of fall detection according to a method according to an embodiment of the invention.

[0032] In an embodiment of the present invention, as shown in FIG. 2 , the first stage is a stage before time T0. In the first stage, a radar point cloud can be continuously output from the radar, and a moving target can be detected based on the radar point cloud. The target's movement may be, for example, a normal movement such as walking, sitting, or standing, or may indicate a fall, or moving or crawling on the ground after a fall. Because the moving target can be detected in the first stage, a fall can be determined based on the radar point cloud information. Furthermore, when the radar point cloud is no longer output from the radar, i.e., when the radar point cloud no longer appears, for example, at time T0, the system transitions to the second stage.

[0033] In an embodiment of the present invention, as shown in FIG. 2 , the second stage is the period from time T0 to time T1. At time T0, the radar point cloud output disappears, and there is no point cloud output until time T1. In the second stage, the detection target transitions from a moving state to a stationary state. At this stage, fall detection can be performed using the radar point cloud and angle FFT information. This is because if the target loses consciousness or becomes extremely weak and unable to move immediately after falling, the radar can only acquire the point cloud from before the target fell to the ground. Because the duration of a fall event is short and there is little point cloud data reflecting the falling motion, it is difficult to detect a fall event using only the radar point cloud. If the target is physiologically active after falling (breathing and heartbeat), the radar will produce relatively large signal fluctuations at the target's position, even if there is no significant body movement. Therefore, by combining the radar point cloud and radar signal fluctuation information, such fall events can be detected.

[0034] In an embodiment of the present invention, as shown in FIG. 2 , the third stage is the period from time T1 to time T2 after the radar point cloud disappears. If a fall event is not detected based on the radar point cloud and angle FFT information in the first and second stages, the process proceeds to the third stage. In a real-world scenario, a slow fall may occur, such as when a target, such as a human body, leans against a wall or other object and slowly slides to the ground. Due to the slow movement of the human body, the radar may not be able to output a valid point cloud, resulting in the failure of the fall detection in the previous two stages. Therefore, in an embodiment of the present invention, a third stage of fall detection is proposed, which determines whether a fall has occurred using only radar angle FFT information.

[0035] In an embodiment of the present invention, a fourth stage may be preferably maintained after the third stage. For example, as shown in FIG. 2, the fourth stage may be maintained for a certain period after the time T2 after the radar point cloud disappears. Because the radar point cloud disappears, the radar coverage area becomes an environment without moving targets. This stationary state may continue for a long time. For example, the target may leave the radar detection area or may fall asleep within the radar coverage area. Since no fall event has been detected after the previous three stages, it is pointless to continue fall detection using radar data at this point. On the other hand, detection over a long period of time may result in false detection due to noise. Therefore, in the fourth stage, if no valid point cloud is output from the radar after the time T2 and no fall event has been detected, fall detection is stopped. The above operation avoids long-term invalid detection, reduces false detection due to noise, and saves computational resources.

[0036] In the above embodiment, if the radar point cloud reappears in the second to fourth stages, i.e., if the radar starts to output a valid point cloud, the system transitions to the first stage, i.e., falls are detected based on the radar point cloud information. If the radar point cloud gradually decreases and disappears, the system returns to the second stage.

[0037] In the above examples, the time lengths of the second, third, and fourth stages are not limited. In one preferred embodiment, the time length of the third stage is longer than the time length of the second stage. For example, the time length of the second stage is set to 2 minutes, and the time length of the third stage is set to 10 minutes. In the third stage, fall detection is performed based only on radar angle FFT information, so it is necessary to accumulate data for a longer period of time to obtain reliable detection results. Therefore, by setting the time length of the third stage longer than that of the second stage, the reliability of the detection results can be improved.

[0038] The following describes each stage of fall detection.

[0039] In the first stage, the radar continues to output point clouds and can detect moving targets. Figure 3 is a schematic diagram of an example of fall detection based on radar point clouds in the first stage. As shown in Figure 3, it includes the following steps:

[0040] Step 310: Filter the radar point cloud for each frame.

[0041] Step 320: Save the filtered radar point cloud in a point cloud information list.

[0042] Step 330: Perform fall detection based on the point cloud in the point cloud information list.

[0043] Note that the above-described Figure 3 merely exemplifies an embodiment of the present invention, and the present invention is not limited thereto. For example, the execution order of each step may be adjusted as appropriate, other steps may be added, or some steps may be deleted. Those skilled in the art may make appropriate modifications based on the above content, and the present invention is not limited to the description of the above-described Figure 3.

[0044] In the above embodiment, first, point cloud information is collected, and then fall detection is performed based on the collected point cloud information.

[0045] When collecting point cloud information, for the radar data of each frame, first, noise in the point cloud is removed by filtering, and then the filtered point cloud is saved in the point cloud information list.

[0046] Normally, even when the reflection intensity of the radar signal of a stationary object is high, a point cloud is generated. However, the Doppler velocity of the point cloud of a stationary object is zero or small. In order to eliminate the influence of the point cloud of a stationary object on the fall judgment, in step 310, the point cloud with a Doppler velocity smaller than V (referred to as the first threshold) may be removed by filtering, that is, the point cloud with |v i |<V may be removed by filtering.

[0047] Also, in step 310, noise in the radar point cloud may be filtered. For example, the point cloud of each frame is clustered using a clustering method, the point cloud belonging to a certain cluster that has been successfully clustered is retained, and the point cloud that does not belong to any cluster and has failed in clustering is removed. The condition for successful clustering is, for example, that the number of point clouds in the cluster is n c (referred to as the second threshold) and larger, and the minimum distance from each point in the point cloud to other point clouds in the same cluster is d c (referred to as the third threshold) and smaller.

[0048] In the above embodiment, after the clustering process, the point cloud is classified into the point cloud belonging to a certain cluster that has been successfully clustered and the point cloud that does not belong to any cluster and has failed in clustering. The former is a valid point cloud, and the latter is noise. Therefore, by clustering the radar point cloud of each frame, the point cloud can be classified into multiple clusters. The number of point clouds in each cluster is n c and larger, and the minimum distance from each point in the point cloud to other point clouds in the same cluster is d c and smaller.

[0049] In the above embodiment, this process may be implemented using a conventional clustering algorithm, such as the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm, although the present invention is not limited thereto.

[0050] In step 320, the point cloud information list is j ,1≦j≦N L}, where P' j is the effective point cloud in the radar data of the jth frame, and N L is the total number of frames.

[0051] In one preferred embodiment, the point cloud information list includes the points closest to the current time point, up to F L The valid point cloud data of frames is stored, that is, N L ≦F L and point cloud data from before that point has been discarded. Also, the number of valid point clouds for each frame in the point cloud information list is greater than zero. If the number of valid point clouds for a frame is 0, the radar data for that frame is ignored from the point cloud information list and does not affect the number of data frames in the point cloud information list. For example, suppose there are currently 50 frames of radar data, and the number of valid point clouds for radar data in 45 of those frames is greater than 0, then the number of data frames stored in the point cloud information list is 45, i.e., N L =45.

[0052] In the above embodiment, the number of data frames in the point cloud information list is F L If the point cloud information list is equal to the point cloud information list, it is determined whether a fall event has occurred.

[0053] For example, the effective point group P' of each frame in the point group information list jIf the number of frames in which the average height z'j of the point cloud in the point cloud information list is smaller than Z0 (referred to as the fourth threshold) is greater than M1 (referred to as the fifth threshold) and the average height z'j of the point cloud of all frames in the point cloud information list is smaller than Z1 (referred to as the sixth threshold), it is determined that a fall event has occurred; otherwise, it is determined that a fall event has not occurred.

[0054] The above is just an example. L There is no restriction on N L That is, the point cloud information list may include only the nearest points from the current point to the maximum of N points. L The valid point cloud data for this frame is retained, and point cloud data from previous points in time is discarded. If the number of valid point clouds for a frame among the valid point cloud data for the frames closest to the current point in time is 0, the radar data for that frame is ignored from the point cloud information list and does not affect the number of data frames in the point cloud information list.

[0055] In the above embodiment, when the radar continuously generates point cloud data and new point cloud information is added to the point cloud information list, the above operation may be repeated to detect a fall event based on the valid point cloud information in the point cloud information list. The first stage detection method is mainly used to detect a situation where the target moves or crawls on the ground after falling.

[0056] In the above embodiment, if a fall event is detected in the first stage, the detection is stopped, and if a fall event is not detected, the process proceeds to the second stage.

[0057] In the above embodiment, a method for detecting a fall based on radar point cloud information has been described as an example, but when actually implementing the method, other methods may be obtained by appropriately modifying the above method, and the description thereof will be omitted here.

[0058] In the second stage, the detection target transitions from a moving state to a stationary state, and in this stage, fall detection may be performed using the radar point cloud and angle FFT information. Figure 4 is a schematic diagram of an example of fall detection based on the radar point cloud information and radar angle FFT information in the second stage. As shown in Figure 4, it includes the following steps.

[0059] Step 410: Valid point cloud data for N1 frames is obtained from the point cloud information list.

[0060] Step 420: Clustering processing is performed on the valid point cloud data of the N1 frames.

[0061] Step 430: Detect the position of the stationary target based on the radar angle FFT information.

[0062] Step 440: Determine whether a rollover event occurs based on the valid point cloud data of the N1 frames and the position of the stationary target.

[0063] Note that the above-described FIG. 4 merely exemplifies an embodiment of the present invention, and the present invention is not limited thereto. For example, the execution order of each step may be adjusted as appropriate, other steps may be added, or some steps may be deleted. Those skilled in the art may make appropriate modifications based on the above content, and are not limited to the description of the above-described FIG. 4.

[0064] In step 410, N1 frames of valid point cloud data are obtained from the point cloud information list. If the number of frames of valid point cloud data in the point cloud information list is smaller than N1, the target's activity time and movement amplitude are small, the possibility of a fall is considered low, and transition to subsequent fall detection in the second stage is unnecessary. If not, subsequent fall detection needs to be performed. The present invention is not limited to the value of N1, which may be determined empirically.

[0065] In step 410, statistics may be taken on the information of valid point clouds of N1 frames for subsequent fall detection, where the number of all point clouds is n1 and the center position of all point clouds is P c =(x c ,y c ,z c ), and the number of points whose height is less than Z2 (called the sixth threshold) is n z2 and the number of points whose height is less than Z3 (called the eighth threshold) is n z3 and the number of points whose height is greater than Z4 (called the 10th threshold) is n z4 The proportion of points whose height is smaller than the sixth threshold Z2 is r z2 =n z2 / n1, and the proportion of points whose height is smaller than the eighth threshold Z3 is r z3 =n z3 / n1.

[0066] In step 420, the valid point cloud data of N1 frames are integrated and clustered to obtain c point cloud clusters. In the above embodiment, the clustering algorithm is not limited, and the clustering may be realized using, for example, the DBSCAN algorithm.

[0067] If the number of point cloud clusters obtained by the above clustering process is greater than 1, i.e., c>1, there are multiple targets within the radar monitoring range, and the possibility of a fall resulting in serious injury is low. Therefore, there is no need to perform the second stage of fall detection; it is determined to be a safe scene and detection is terminated.

[0068] If the number of point cloud clusters obtained by the above clustering process is one or less, the position of the stationary target is detected based on the radar angle FFT information.

[0069] In step 430, the radar angle FFT information can be used to estimate the position of stationary targets (eg, human bodies) since physiological activity of the human body affects the radar signal.

[0070] In one preferred embodiment, the position of a stationary target is estimated based on the signal fluctuation of the radar angle FFT. The fluctuation of the angle FFT amplitude at each spatial position within a certain period is calculated, and the fluctuation may be expressed using the standard deviation of the angle FFT amplitude. If the fluctuation value of the radar signal is greater than a threshold, it is determined that a target is present within the radar monitoring range, and the position where the radar signal fluctuation value is maximum is the position of the stationary target.

[0071] In another preferred embodiment, the signal fluctuation and frequency spectrum characteristics of the radar angle FFT are used to estimate the position of a stationary target. For details, reference may be made to the related art, and the description thereof will be omitted here.

[0072] In the above embodiment, the radar continues to output angle FFT information as long as the monitored target is stationary, so the position of the stationary target can be detected multiple times.

[0073] In the above example, the position of the stationary target detected based on the angle FFT information after the radar point cloud disappears is P s ={(x s,i ,y s,i ,z s,i ), 1≦i≦n s}, where n s is the number of detections.

[0074] First, as shown in the following equation (1), P s The position of the stationary target and the center position P of the effective point group c Distance d i Calculate.

[0075]

number

[0076] < In the above embodiment, r< z2 (the ratio of the point cloud with a height smaller than the 6th threshold Z2) is larger than R2 (referred to as the 7th threshold), r< z3 (the ratio of the point cloud with a height smaller than the 8th threshold Z3) is larger than the threshold R3 (referred to as the 9th threshold), n< z4 (the number of point clouds with a height larger than the 10th threshold Z4) is smaller than N< z4 (the 11th threshold), and when r< z5 (the number of detection results satisfying the above conditions among the multiple detection results of the position of the stationary target) is larger than R4 (referred to as the 14th threshold), it is determined that a fall event has occurred, and otherwise, it is determined that no fall has occurred. < <

[0077] < In the above embodiment, in the second stage, when a fall event is detected, the detection is stopped, and when no fall event is detected, it proceeds to the third stage. < <

[0078] < In the above embodiment, in the second stage, when a radar point cloud appears, it proceeds to the first stage and performs fall detection based on the radar point cloud information. Since the specific detection method has already been described, the description is omitted here. < <

[0079] < In the above embodiment, a method for detecting a fall based on radar point cloud information and radar angle FFT information has been described as an example. However, in the course of specific implementation, other methods may be obtained by appropriately modifying the above method, and the description thereof will be omitted here.

[0080] In the third stage, whether a fall has occurred is determined using only the radar angle FFT information. Figure 5 is a schematic diagram of an example of fall detection based on the radar angle FFT information in the third stage. As shown in Figure 5, it includes the following steps:

[0081] Step 510: Detect the position of the stationary target based on the radar angle FFT information.

[0082] Step 520: Determine whether a rollover event has occurred based on the position of the stationary target.

[0083] In the above embodiment, as explained above, the radar angle FFT information can be used to estimate the position of stationary targets.

[0084] In the above example, the position of the stationary target detected based on the angle FFT information after the radar point cloud disappears is P s ={(x s,i ,y s,i ,z s,i ), 1≦i≦n s}, and n s is the number of detections.

[0085] In one preferred embodiment, before time T2, n s N s (referred to as the 15th threshold) and P s z in (a set of stationary target positions obtained by multiple detections) s,i If the ratio of (the height of the stationary target position) being smaller than Z5 (the 13th threshold) is greater than R5 (referred to as the 16th threshold), it is determined that a fall event has occurred; otherwise, it is determined that a fall event has not occurred.

[0086] In the above embodiment, similar to the second stage of detection, if a radar point cloud appears in the third stage, the process shifts to the first stage, and fall detection is performed based on the radar point cloud information. The specific detection method has already been explained, so its explanation will be omitted here.

[0087] In the above embodiment, only the method of detecting a fall based on radar angle FFT information has been described as an example. However, in the course of specific implementation, other methods may be obtained by appropriately modifying the above method, and the description thereof will be omitted here.

[0088] In the embodiment of the present invention, a fourth step may be preferably added, as shown in Figure 2. As described above, in the fourth step, if no valid point cloud is output from the radar after time T2 and no fall event is detected, the fall detection is stopped. This operation can avoid long-term invalid detection, reduce false detections due to noise, and save computational resources.

[0089] The above-described embodiments are merely illustrative of the embodiments of the present invention, and the present invention is not limited thereto. Appropriate modifications may be made based on the above-described various embodiments. For example, the above-described embodiments may be used alone, or one or more of the above-described embodiments may be used in combination.

[0090] 6 is a schematic diagram of an example of a fall detection process of the method according to an embodiment of the present invention. As shown in FIG. 6, the process includes the following steps:

[0091] Step 610: Filter the point cloud from the static target, |v i | <Vである。

[0092] Step 620: Filter out noise in the point cloud through clustering to obtain a valid point cloud.

[0093] Step 630: Convert the valid point group into the point group information list L={P' j, 1 ≤ j ≤ N L Save it in {}.

[0094] Step 640: Calculate the average height z’ of the point cloud of each frame in L j Calculate it.

[0095] Step 650: N L = F L If the number of frames where z’ < Z0 is greater than M1 and for all frames z’ < Z1, it is detected that a fall has occurred, end the detection; otherwise, execute Step 660. j If the number of frames where z’ < Z0 is greater than M1 and for all frames z’ < Z1, it is detected that a fall has occurred, end the detection; otherwise, execute Step 660. j If the number of frames where z’ < Z0 is greater than M1 and for all frames z’ < Z1, it is detected that a fall has occurred, end the detection; otherwise, execute Step 660.

[0096] Step 660: Obtain the point cloud of N1 frames from L during the period between time point T0 and time point T1 (the second stage). [[ID=2`6]]

[0097] Step 670: Cluster all the point clouds of N1 frames to obtain c clusters.

[0098] Step 680: If c > 1, a safe scene is detected, end the detection; otherwise, execute Step 690.

[0099] Step 6`90: r z2 = n z2 / n1 and r z3 = n z3 / n1 are calculated. [[ID=;45]]

[0100] Step 6100: Use the angular FFT information to perform the detection of static targets and obtain P s = {(x s,i , y s,i , z s,i ), 1 ≤ i ≤ n s} <;

[0101] Step 6110: Calculate the ratio r z5 = n z5 / n s Calculate it.

[0102] Step 6120:r z2 >R2 and r z3 >R3, and n z4 <N z4 and r z5 If >R4, detect that a fall has occurred, otherwise execute step 6130.

[0103] Step 6130: Between time T1 and time T2 (third stage), n s >N s and P s ni z s,i If the ratio of Z5 being smaller than Z5 is greater than R5 , it is detected that a fall has occurred and the detection is terminated; otherwise, step 6140 is executed.

[0104] Step 6140: After time T2 (fourth stage), fall detection is stopped.

[0105] Note that the above-described Figure 6 merely exemplifies an embodiment of the present invention, and the present invention is not limited thereto. For example, the execution order of each step may be adjusted as appropriate, other steps may be added, or some steps may be deleted. Those skilled in the art may make appropriate modifications based on the above content, and are not limited to the description of the above-described Figure 6.

[0106] According to an embodiment of the present invention, the speed and reliability of fall detection can be improved by performing fall detection using radar point cloud information and radar angle FFT information at different stages when a fall may occur.

[0107] <Example 2> The present embodiment provides a fall detection device, the solution principle of which is the same as that of the method of the first embodiment, and therefore the specific implementation thereof may refer to the implementation of the method of the first embodiment, and the same content will not be described again.

[0108] 7 is a schematic diagram of an example of a fall detection device according to an embodiment of the present invention. As shown in FIG. 7, a fall detection device 700 according to an embodiment of the present invention includes the following components.

[0109] The detection unit 710 performs fall detection based on radar point cloud information in a first stage, and if a fall event is not detected, proceeds to a second stage, in which it performs fall detection based on radar point cloud information and radar angle FFT information in the second stage, and if a fall event is not detected, proceeds to a third stage, in which it performs fall detection based on radar angle FFT information.

[0110] In some embodiments, in the first stage, if a fall event is detected, detection is stopped (YES in step 650 shown in FIG. 6 ); in the second stage, if a fall event is detected (YES in step 6120 shown in FIG. 6 ) or a safe scene is detected (YES in step 680 shown in FIG. 6 ), detection is stopped; in the third stage, if a fall event is detected (YES in step 6130 shown in FIG. 6 ), detection is stopped; and if a fall event is not detected (NO in step 6130 shown in FIG. 6 ), it is determined that a fall has not occurred.

[0111] In some embodiments, the first stage is a stage where a radar point cloud is present, and if the radar point cloud disappears, the second stage is entered. The second and third stages are stages where a radar point cloud is not present. The duration of the second stage may be shorter than the duration of the third stage.

[0112] In some embodiments, if a new radar point cloud appears in the second and third stages, the process transitions back to the first stage.

[0113] In some embodiments, in the first stage, when performing fall detection based on radar point cloud information, the detection unit 710 filters the radar point cloud of each frame, stores the filtered radar point cloud in a point cloud information list, and performs fall detection based on the point cloud in the point cloud information list.

[0114] In the above embodiment, filtering the radar point cloud for each frame may include filtering out points with a Doppler velocity less than a first threshold V.

[0115] In the above embodiment, filtering the radar point cloud of each frame may include clustering the point cloud of each frame using a clustering method, retaining point clouds that are successfully clustered and belong to a cluster, and removing point clouds that are unsuccessful in clustering and do not belong to any cluster.

[0116] In the above example, the condition for successful clustering is that the number of points in a cluster exceeds the second threshold n c and the minimum distance from each point in the point cloud to other points in the same cluster is greater than a third threshold d c It may be smaller than

[0117] In the above embodiment, the point cloud information list is L={P' j ,1≦j≦N L}, where P' j is the effective point cloud in the radar data of the jth frame, and N L is the total number of frames.

[0118] In the above embodiment, if the number of frames in which the average height z'j of the point cloud in the point cloud information list is smaller than the fourth threshold Z0 is greater than the fifth threshold M1, and the average height z'j of the point cloud of all frames in the point cloud information list is smaller than the sixth threshold Z1, it is determined that a fall event has occurred; otherwise, it is determined that a fall event has not occurred.

[0119] In some embodiments, in the second stage, when performing fall detection based on radar point cloud information and radar angle FFT information, the detection unit 710 obtains effective point cloud data of N1 frames from the point cloud information list, performs clustering processing on the effective point cloud data of the N1 frames, and if the number of point cloud clusters obtained by the clustering processing is 1 or less, detects the position of a stationary target based on the radar angle FFT information, and determines whether a fall event has occurred based on the effective point cloud data of the N1 frames and the position of the stationary target.

[0120] In the above embodiment, if the number of frames of valid point cloud data in the point cloud information list is smaller than N1, it is determined that a fall has not occurred (detected as not having occurred), and detection ends.

[0121] In the above embodiment, if the number of point cloud clusters is greater than 1, it is determined to be a safe scene and detection is terminated.

[0122] In the above embodiment, if the conditions are met, it is determined that a fall event has occurred, and if not, it is determined that a fall event has not occurred. The proportion r of points whose height is smaller than the sixth threshold Z2 z2 is greater than a seventh threshold R2; The proportion r of points whose height is smaller than the eighth threshold Z3 z3 is greater than a ninth threshold R3; The number of points n whose height is greater than the tenth threshold Z4 z4 is the 11th threshold N z4 is less than A set of stationary target positions P obtained by multiple (at least one) detections s The distance d between the position of the stationary target and the center of the effective point cloud of the N1 frames is i is smaller than the twelfth threshold D1, and the height z s,i is smaller than the 13th threshold Z5 z5 The proportion r z5 is greater than a fourteenth threshold R4.

[0123] In some embodiments, in the third stage, when performing fall detection based on the radar angle FFT information, the detection unit 710 detects the position of a stationary target based on the radar angle FFT information, and determines whether a fall event has occurred based on the position of the stationary target.

[0124] In the above embodiment, the number of detections n s is the 15th threshold N s A set P of stationary target positions that is larger than s Among these, the height z of the stationary target s,i If the ratio of the number of times that the number of times ...

[0125] In some embodiments, the detector 710 stops fall detection if it still has not detected a fall event after the third period of time, and preferably returns to the first stage if a radar point cloud appears within a certain time (fourth stage) after stopping fall detection.

[0126] Although the above description only covers the components and modules relevant to the present invention, the present invention is not limited to these. The fall detection device 700 may include other components or modules, and reference may be made to related art for specific details of these components or modules.

[0127] For simplicity, Fig. 7 only exemplifies the connection relationships or signal flows between individual components or modules, but it will be apparent to those skilled in the art that various related technologies, such as bus connections, may also be used. Each of the above components or modules may be realized by hardware equipment, such as a processor, a memory, etc. However, the embodiments of the present invention are not limited thereto.

[0128] The above-mentioned embodiments are merely examples of the present invention, but the present invention is not limited to these and can be modified as appropriate in addition to the above-mentioned embodiments. For example, each of the above-mentioned embodiments may be used alone, or one or more of the above-mentioned embodiments may be combined.

[0129] According to an embodiment of the present invention, the speed and reliability of fall detection can be improved by performing fall detection using radar point cloud information and radar angle FFT information at different stages when a fall may occur.

[0130] Example 3 An embodiment of the present invention provides a computer device including the fall detection device 700 according to Example 2, the contents of which are incorporated herein by reference. The computer device may be, for example, a computer, a server, a workstation, a laptop computer, a smartphone, or the like, but the embodiment of the present invention is not limited thereto.

[0131] 8 is a schematic diagram of an example of a computer device according to an embodiment of the present invention. As shown in FIG. 8, the computer device 800 includes a processor (e.g., a central processing unit (CPU)) 810 and a memory 820. The memory 820 is connected to the processor 810. The memory 820 may store various data and may further store an information processing program 821. The program 821 is executed under the control of the processor 810.

[0132] In some embodiments, the functionality of the fall detection device 700 may be integrated into the processor 810. Here, the processor 810 may be configured to implement the fall detection method described in Example 1.

[0133] In some embodiments, the fall detection device 700 may be disposed with the processor 810, for example, the fall detection device 700 may be a chip connected to the processor 810 and configured to realize the functions of the fall detection device 700 under the control of the processor 810.

[0134] 8, the computer device 800 may further include an input / output (I / O) device 830, a display 840, etc. Here, the functions of the above components are the same as those of the prior art, and therefore, description thereof will be omitted here. Note that the computer device 800 does not need to include all of the components shown in FIG. 8. Furthermore, the computer device 800 may include components not shown in FIG. 8, and reference may be made to related art.

[0135] An embodiment of the present invention provides a computer-readable program that, when executed in a fall detection device, causes a living body position detection device to execute the method described in embodiment 1.

[0136] An embodiment of the present invention further provides a storage medium storing a computer-readable program for causing a fall detection device to execute the method according to the first embodiment.

[0137] The above-described apparatus and method of the present invention may be realized by hardware or a combination of hardware and software. The present invention relates to a computer-readable program that, when executed by a logic unit, causes the logic unit to realize the above-described apparatus or components, or to implement the above-described various methods or steps. The present invention also relates to a storage medium for storing the above-described program, such as a hard disk, magnetic disk, optical disk, DVD, flash memory, etc.

[0138] The methods / apparatuses described with reference to the embodiments of the present invention may be implemented in hardware, software modules executed by a processor, or a combination of both. For example, one or more of the functional block diagrams shown in the drawings, or one or more combinations of the functional block diagrams, may correspond to software modules in a computer program flow or hardware modules. These software modules may correspond to steps shown in the drawings. These hardware modules may be implemented by implementing these software modules in hardware, for example, using a field programmable gate array (FPGA).

[0139] The software module may be located in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, mobile hard disk, CD-ROM, or any other form of storage medium known to those skilled in the art. The storage medium may be connected to the processor so that the processor reads information from or writes information to the storage medium, or the storage medium may be a component of the processor. The processor and the storage medium are located in an ASIC. The software module may be stored in the memory of the mobile terminal or in a memory card inserted into the mobile terminal. For example, if a device (e.g., a mobile terminal) uses a relatively large-capacity MEGA-SIM card or a large-capacity flash memory device, the software module may be stored in the MEGA-SIM card or the large-capacity flash memory device.

[0140] One or more of the functional blocks and / or one or more combinations of functional blocks illustrated in the figures may be implemented with a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, or any suitable combination thereof to perform the functions described herein. One or more of the functional blocks and / or one or more combinations of functional blocks illustrated in the figures may be implemented with, for example, a combination of computing devices, such as a combination of a DSP and a microprocessor, a combination of multiple microprocessors, one or more microprocessors in combination with a DSP communication, or any other configuration.

[0141] Although the present invention has been described above with reference to specific embodiments, the above description is merely illustrative and does not limit the scope of protection of the present invention. Various modifications and changes may be made to the present invention without departing from the spirit and principles of the present invention, and these modifications and changes also fall within the scope of the present invention.

[0142] Furthermore, the following supplementary notes are disclosed regarding the embodiments including the above-mentioned examples. (Appendix 1) A fall detection method, comprising: In the first stage, performing fall detection based on radar point cloud information, and if a fall event is not detected, proceeding to a second stage; In the second stage, detecting a fall based on the radar point cloud information and the radar angle FFT information, and if a fall event is not detected, proceeding to a third stage; and in a third stage, performing fall detection based on the radar angle FFT information. (Appendix 2) The first stage is a stage where a radar point cloud exists, and when the radar point cloud disappears, the process proceeds to a second stage; 2. The method of claim 1, wherein the second and third stages are stages in which no radar point cloud exists. (Appendix 3) 3. The method of claim 2, wherein the duration of the second stage is shorter than the duration of the third stage. (Appendix 4) In the first stage, the step of detecting a fall based on radar point cloud information includes: filtering the radar point cloud for each frame; storing the filtered radar point cloud in a point cloud information list; performing fall detection based on the point cloud in the point cloud information list. (Appendix 5) The point cloud information list is L={P' j ,1≦j≦N L}, where P' j is the effective point cloud in the radar data of the jth frame, and N L is the total number of frames. (Appendix 6) The method described in Appendix 4, wherein if the number of frames in which the average height z'j of the point cloud in the point cloud information list is smaller than a fourth threshold Z0 is greater than a fifth threshold M1 and the average height z'j of the point cloud of all frames in the point cloud information list is smaller than a sixth threshold Z1, it is determined that a fall event has occurred; otherwise, it is determined that a fall event has not occurred. (Appendix 7) In the second stage, the step of detecting a fall based on the radar point cloud information and the radar angle FFT information includes: A step of acquiring valid point cloud data of N1 frames from the point cloud information list; performing clustering processing on the valid point cloud data of the N1 frames; If the number of point cloud clusters obtained by the clustering process is one or less, detecting the position of a stationary target based on radar angle FFT information; A method according to any one of appendices 1 to 3, comprising determining whether a rollover event has occurred based on the valid point cloud data of the N1 frames and the position of the stationary target. (Appendix 8) If the number of frames of valid point cloud data in the point cloud information list is smaller than N1, it is determined that a fall has not occurred and detection is terminated; and / or 8. The method of claim 7, wherein if the number of point cloud clusters is greater than 1, the scene is determined to be safe and detection is terminated. (Appendix 9) In the third stage, the step of detecting a fall based on radar angle FFT information includes: Detecting a stationary target position based on the radar angle FFT information; determining whether a rollover event has occurred based on the position of the stationary target; Number of detections n s is the 15th threshold N s and the position set P of the stationary target obtained by multiple detections is larger than s , the height z of the stationary target s,i is smaller than a thirteenth threshold Z5 is greater than a sixteenth threshold R5, it is determined that a fall event has occurred, and if not, it is determined that a fall event has not occurred. (Appendix 10) 10. The method of any of claims 1 to 9, further comprising the step of stopping fall detection if a fall event is still not detected after the third stage.

Claims

1. A fall detection device, comprising: a detection unit that, in a first stage, performs fall detection based on radar point cloud information, and if a fall event is not detected, transitions to a second stage, and in the second stage, performs fall detection based on the radar point cloud information and radar angle FFT information, and if a fall event is not detected, transitions to a third stage, and in the third stage, performs fall detection based on the radar angle FFT information.

2. In the first stage, the detection unit stops detection when a fall event is detected; In the second stage, the detection unit stops detection when a fall event is detected; The device according to claim 1 , wherein in the third stage, the detection unit stops detection if a fall event is detected, and determines that a fall has not occurred if a fall event is not detected.

3. If a new radar point cloud appears in the second stage, the process proceeds to the first stage; The apparatus of claim 1 , wherein if a new radar point cloud appears in the third stage, the apparatus transitions to the first stage.

4. When performing fall detection based on radar point cloud information in the first stage, the detection unit filtering the radar point cloud for each frame to filter out points having a Doppler velocity less than a first threshold V and / or noise within the point cloud; Save the filtered radar point cloud in the point cloud information list, The device of claim 1 , wherein fall detection is performed based on the point clouds in the point cloud information list.

5. Filtering out noise in the point cloud is clustering the point clouds of each frame using a clustering method, retaining point clouds that are successfully clustered and belong to a cluster, and removing point clouds that are unsuccessful in clustering and do not belong to any cluster; The condition for the clustering to be successful is that the number of points in the cluster is equal to or exceeds a second threshold n ｃ and the minimum distance from each point in the point cloud to other points in the same cluster is greater than a third threshold d ｃ 5. The apparatus of claim 4, wherein the distance is less than .

6. In the second stage, when performing fall detection based on radar point cloud information and radar angle FFT information, the detection unit From the point cloud information list １ Obtain valid point cloud data for frames, The N １ Clustering is performed on the valid point cloud data of the frames, If the number of point cloud clusters obtained by the clustering process is one or less, detecting the position of a stationary target based on radar angle FFT information; The N １ The apparatus of claim 1 , further comprising: determining whether a rollover event has occurred based on the valid point cloud data of the frames and the positions of the stationary targets.

7. The number of frames of valid point cloud data in the point cloud information list is N １ The device according to claim 6 , wherein if the difference is smaller than the predetermined value, it is determined that a fall has not occurred, and the detection unit terminates detection.

8. The apparatus according to claim 6 , wherein if the number of point cloud clusters is greater than one, the scene is determined to be a safe scene, and the detection unit terminates the detection.

9. If the condition is met, it is determined that a fall event has occurred, and if not, it is determined that a fall event has not occurred, and the condition is: The height is the sixth threshold Z ２ The proportion of points smaller than r ｚ２ is the seventh threshold R ２ That is greater than The height is the eighth threshold Z ３ The proportion of points smaller than r ｚ３ is the ninth threshold R ３ That is greater than The height is the 10th threshold Z ４ The number of points n that is greater than ｚ４ is the eleventh threshold N ｚ４ is less than A set of stationary target positions P obtained by multiple detections ｓ , the position of the stationary target and the N １ Distance d from the center of the effective point group of this frame ｉ is the 12th threshold D １ and the height z of the stationary target ｓ，ｉ is the 13th threshold Z ５ A number n that satisfies the condition that it is smaller than ｚ５ The proportion r ｚ５ is the fourteenth threshold R ４ The apparatus of claim 6 , wherein the first and second inputs are greater than or equal to 100 kJ / s.

10. When performing fall detection based on radar angle FFT information in the third stage, the detection unit Detecting the position of a stationary target based on the radar angle FFT information; The device of claim 1 , further comprising: determining whether a rollover event has occurred based on the position of the stationary target.

Images