CN112036269A - Fall detection method and device, computer equipment and storage medium - Google Patents
Fall detection method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN112036269A CN112036269A CN202010824761.4A CN202010824761A CN112036269A CN 112036269 A CN112036269 A CN 112036269A CN 202010824761 A CN202010824761 A CN 202010824761A CN 112036269 A CN112036269 A CN 112036269A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- human body
- space
- detected
- height
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 75
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000012360 testing method Methods 0.000 claims description 85
- 238000004590 computer program Methods 0.000 claims description 25
- 238000010606 normalization Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000005070 sampling Methods 0.000 description 18
- 230000006399 behavior Effects 0.000 description 8
- 238000013135 deep learning Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000012795 verification Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000005452 bending Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000009429 distress Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000002366 time-of-flight method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Abstract
The application relates to a fall detection method, apparatus, computer device and storage medium. The method comprises the following steps: acquiring a space point cloud of a space to be detected; identifying a human body point cloud set in the spatial point cloud; determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected; if the height of the human body point cloud is lower than a preset height threshold value, judging that the human body point cloud is a first target point cloud; and judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the space to be detected has a falling condition according to the judgment result. By adopting the method, the accuracy of the fall detection result can be improved.
Description
Technical Field
The present application relates to the field of digital image processing technologies, and in particular, to a method and an apparatus for fall detection based on depth point cloud data, a computer device, and a storage medium.
Background
The falling behavior of the human body can be detected timely and accurately, a distress signal can be sent timely when the human body falls, and rescue can be conducted timely, so that the falling behavior detection device has an important significance in ensuring the life safety of people. The traditional fall detection equipment comprises wearable detection equipment and RGB (red, green and blue) camera detection equipment, wherein the wearable detection equipment needs to be worn by a user for a long time, discomfort is easily caused, and the user generally has higher requirement on the safety of a product; the RGB camera detection equipment is difficult to be applied to places such as bathrooms and bedrooms where privacy needs to be maintained, has high requirements on illumination and is not suitable for night detection.
The depth camera detection equipment can detect the falling behavior of a human body by shooting the depth image, does not infringe the privacy, is not influenced by illumination, and can realize a better detection effect at night. At present, a method for fall detection by using a depth image usually judges whether a fall occurs according to human body joints in the depth image, however, the detection method is easily affected by the body type change of a detected person, and the accuracy of a detection result is low.
Therefore, the current fall detection technology has the problem of low accuracy of detection results.
Disclosure of Invention
In view of the above, it is necessary to provide a fall detection method, an apparatus, a computer device, and a storage medium capable of improving accuracy of detection results in view of the above technical problems.
A fall detection method, the method comprising:
acquiring a space point cloud of a space to be detected;
identifying a human body point cloud set in the spatial point cloud;
determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected;
if the height of the human body point cloud is lower than a preset height threshold value, judging that the human body point cloud is a first target point cloud;
and judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the space to be detected has a falling condition according to the judgment result.
In one embodiment, the identifying a human body point cloud set in the spatial point cloud includes:
identifying a ground extent in the spatial point cloud;
performing clustering classification on the spatial point cloud according to the ground range to obtain an object point cloud set in the space to be detected;
and identifying the object point cloud set to obtain a human body point cloud set in the space to be detected.
In one embodiment, the identifying the ground extent in the spatial point cloud comprises:
randomly selecting a test point cloud in the space point cloud;
determining a test plane corresponding to the test point cloud;
marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds;
judging whether the number of the second target point clouds exceeds a preset number threshold value;
if not, returning to the step of randomly selecting the test point cloud in the space point cloud;
and if so, obtaining the ground range in the spatial point cloud according to the test plane.
In one embodiment, the determining the test plane corresponding to the test point cloud includes:
substituting the point cloud data of the test point cloud into an initial space plane formula to obtain a formula coefficient of the initial space plane formula;
determining a target space plane formula according to the formula coefficient;
and obtaining the test plane according to the target space plane formula.
In one embodiment, the obtaining the human body point cloud set in the space to be detected by identifying the object point cloud set includes:
counting the number of the point clouds in the object point cloud set;
if the number of the point clouds in the object point cloud set conforms to a preset number range, taking the object point cloud set as a candidate point cloud set;
judging whether the candidate point cloud set is a human body point cloud set or not through a pre-training model;
and if so, determining the candidate point cloud set as the human body point cloud set.
In one embodiment, the determining the height of each human point cloud in the human point cloud set according to the ground position in the space to be detected includes:
calculating the distance between the human body point cloud and the ground position to obtain the absolute point cloud height of the human body point cloud;
and carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud.
In one embodiment, the determining whether a falling condition exists in the space to be detected according to the determination result includes:
if the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold, acquiring the duration of the condition that the ratio of the first target point cloud in the human body point cloud set exceeds the preset ratio threshold;
and if the duration exceeds a preset duration threshold, determining that the falling condition exists in the space to be detected.
A fall detection apparatus, the apparatus comprising:
the acquisition module is used for acquiring a spatial point cloud of a space to be detected;
the human body point cloud identification module is used for identifying a human body point cloud set in the space point cloud;
the height calculation module is used for determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected;
the target point cloud identification module is used for judging that the human body point cloud is a first target point cloud if the height of the human body point cloud is lower than a preset height threshold;
and the falling detection module is used for judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value or not and determining whether a falling condition exists in the space to be detected according to a judgment result.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring a space point cloud of a space to be detected;
identifying a human body point cloud set in the spatial point cloud;
determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected;
if the height of the human body point cloud is lower than a preset height threshold value, judging that the human body point cloud is a first target point cloud;
and judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the space to be detected has a falling condition according to the judgment result.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring a space point cloud of a space to be detected;
identifying a human body point cloud set in the spatial point cloud;
determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected;
if the height of the human body point cloud is lower than a preset height threshold value, judging that the human body point cloud is a first target point cloud;
and judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the space to be detected has a falling condition according to the judgment result.
According to the method, the device, the computer equipment and the storage medium for detecting falling, the space point cloud of the space to be detected is obtained, the human body point cloud set in the space point cloud is identified, and falling detection can be carried out on the human body point cloud set in the space to be detected; the height of each human body point cloud in the human body point cloud set is determined according to the ground position in the space to be detected, falling detection can be carried out according to the height of the human body point cloud in the human body point cloud set, and detection accuracy is improved; if the height of the human body point cloud is lower than a preset height threshold, the human body point cloud is judged to be a first target point cloud, falling detection can be further carried out according to the first target point cloud lower than the preset height threshold, whether the proportion of the first target point cloud in the human body point cloud set exceeds a preset proportion threshold or not is judged, whether a falling condition exists in a space to be detected or not is determined according to a judgment result, the falling condition exists in the space to be detected can be judged when the human body point cloud in the human body point cloud set is generally close to the ground, and the accuracy of the detection result is improved.
Drawings
Fig. 1 is a diagram of an application environment of a fall detection method in an embodiment;
fig. 2 is a schematic flow chart of a fall detection method in an embodiment;
fig. 3 is a schematic flow chart of a fall detection method in another embodiment;
fig. 4 is a block diagram of the structure of a fall detection apparatus in one embodiment;
FIG. 5 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The fall detection method provided by the present application can be applied to the application environment shown in fig. 1. Wherein the depth camera 102 communicates with the server 104 via a wired or wireless link. The depth camera 102 may be, but not limited to, various cameras for acquiring depth point cloud data in space by structured light, binocular vision, and time-of-flight methods, and the server 104 may be implemented by an independent server or a server cluster composed of a plurality of servers.
In one embodiment, as shown in fig. 2, a fall detection method is provided, which is exemplified by the application of the method to the server 104 in fig. 1, and includes the following steps:
step S210, obtaining a spatial point cloud of a space to be detected.
The space to be detected is a space range shot by the depth camera 102, and the space point cloud is space point cloud data collected by the depth camera 102.
In a specific implementation, the server 104 may obtain a spatial point cloud of the space to be detected through the depth camera 102. One or more depth cameras 102 can be installed in a room, a certain space range in the room is shot in real time through the depth cameras 102 to obtain multiple frames of continuous depth images, the space range shot by the depth cameras 102 is a space to be detected, the depth image data shot by the depth cameras is space point cloud data of the space to be detected, after the space point cloud data are collected, the depth cameras 102 can transmit the space point cloud data to a server 104, the server 104 receives the space point cloud data, the space point cloud data corresponding to each frame of depth image are processed, and falling conditions in the shooting range of the depth cameras 102 are detected. The spatial point cloud data may be three-dimensional coordinate data of all objects in the space to be detected. For example, the depth camera 102 may use its own position as a coordinate origin to acquire three-dimensional coordinate data of all objects in a shooting range in real time according to a preset coordinate axis, so as to obtain a three-dimensional coordinate data set { (10,20,10), (25,35,50), …, (15,20,40) }, where each three-dimensional coordinate data in the three-dimensional coordinate data set may be used as a spatial point cloud data.
And step S220, identifying a human body point cloud set in the space point cloud.
The human body point cloud is point cloud data corresponding to a human body, and the human body point cloud set is a set formed by the human body point clouds in the space point cloud.
In a specific implementation, the server 104 may first identify a ground range in the spatial point cloud, then perform cluster classification on the spatial point cloud data in the ground range, where a group of point cloud data obtained by the cluster classification may be marked as a three-dimensional object, and perform cluster classification on all spatial point cloud data to obtain a point cloud data set corresponding to each object in the ground range, that is, an object point cloud set, and then identify the object point cloud set, where if an identification result is a point cloud data set corresponding to a human body, the object point cloud set may be marked as a human body point cloud set, and each three-dimensional coordinate data in the human body point cloud set is point cloud data corresponding to the human body.
In a specific embodiment, the ground range in the spatial point cloud may be identified by a spatial plane formula, the spatial plane formula may be represented as ax + by + cz + d being 0, three points are randomly selected from the spatial point cloud and substituted into the spatial plane formula, so that plane coefficients a, b, c, and d of the spatial plane formula may be determined, a specific expression of the spatial plane formula may be determined according to the plane coefficients, an exact spatial plane may be obtained, in order to further determine whether the spatial plane is the ground, three points may be repeatedly selected by probability iteration for verification, and if the number of point clouds in the spatial plane exceeds a preset threshold within a preset number of iterations, it may be determined that the current spatial plane is the ground.
For example, the number of iterations is initially set to 0, and three points (x) are randomly selected from the spatial point cloud1,y1,z1)，(x2,y2,z2)，(x3,y3,z3) Substituting the space plane formula into the space plane formula to establish a joint equation set, obtaining a plane coefficient of a being 5, b being 8, c being-10, d being-6, further determining the space plane formula of 5x +8y-10z-6 being 0, determining a unique space plane according to the formula, calculating the distance between the space point cloud data and the space plane, if the distance is less than a certain threshold (for example, 3 cm), determining that the current space point cloud data is the point cloud data in the space plane, if the proportion of the number of the point cloud data in the space plane to the number of all the space point cloud data is less than a certain proportion threshold (for example, 30%), determining that the space plane is not the ground, adding 1 to the iteration number, and selecting three points (x) at random (x is x)4,y4,z4)，(x5,y5,z5)，(x6,y6,z6) Repeating the above process, and under the condition that the iteration number does not exceed a certain threshold (for example, 100 times), if the proportion of the number of the point cloud data in the space plane to the number of all the space point cloud data exceeds 30%, determining that the current space plane is the ground, and the current space plane formula is the ground formula, which can be determined according to the formulaThe ground formula determines a ground range in the space to be detected, for example, a ground range can be formed according to all three-dimensional data which are determined according to a certain sampling interval and accord with the ground formula.
In another embodiment, since the falling position of the human body is usually the ground, the falling detection can be performed based on the ground range. In the ground range, Clustering and classifying all Spatial point cloud data to obtain a point cloud data set corresponding to each object, and specifically, performing DBSCAN (Density-Based Spatial Clustering of Applications with Noise) Clustering and classifying processing on the Spatial point cloud by using an euclidean distance formula to obtain a point cloud data set corresponding to each object in each frame of depth image, where the euclidean distance Clustering and classifying formula may be
Wherein x is1i，x2iTwo point cloud data, and D is the Euclidean distance between the point cloud data. For the point cloud data set corresponding to each object, it may be preliminarily screened according to the size of the point cloud data set, for example, the size of the point cloud set of the human body may be preset to be in the range of [100, 200 [, 200]]If the size of an object point cloud set belongs to the range, the object point cloud set may be a human body point cloud set, the object point cloud set is marked as a candidate point cloud set, then the candidate point cloud set can be identified through deep learning, and whether the object point cloud set is a human body point cloud data set is detected, for example, whether the candidate point cloud set is the human body point cloud data set can be judged through an OpenVINO (a tool kit capable of accelerating high-performance computer vision and deep learning vision application and development speed). The point cloud data set of the human body can correspond to complete three-dimensional data of one person in one frame of depth image, a nonnegative integer number of the point cloud data set of the human body can exist in one frame of depth image, and when the point cloud data sets of the human body are multiple, the falling conditions of multiple human bodies can be detected.
And step S230, determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected.
The ground position is the position of the ground in the space to be detected determined according to a ground formula.
In a specific implementation, after determining that the current spatial plane formula is the ground formula, the server 104 may determine a position of a sampling point in the space to be detected, where the sampling point corresponds to the ground formula, as a ground position, where the sampling point may be three-dimensional sampling data obtained by sampling the space to be detected according to a certain sampling interval, and a position where the three-dimensional sampling data corresponding to the ground formula is located may be a ground position. The height of the human body point cloud can be obtained by calculating the distance between each human body point cloud data in a human body point cloud set and the ground position, and the specific formula can be
Wherein (x)0,y0,z0) Is a human body point cloud data in a human body point cloud set, h is a human body point cloud data (x)0,y0,z0) Of (c) is measured.
In step S240, if the height of the human body point cloud is lower than the preset height threshold, it is determined that the human body point cloud is the first target point cloud.
In a specific implementation, the server 104 may compare the height of the human point cloud with a preset height threshold, and if the height of the point cloud is lower than the height threshold, it may determine that the corresponding human point cloud is the first target point cloud, for example, the height threshold may be set to be 30 centimeters, and for a human point cloud set, it may determine that the human point cloud with the height lower than 30 centimeters in the set is the first target point cloud, and the server 104 may further mark the first target point cloud.
And step S250, judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the falling condition exists in the space to be detected according to the judgment result.
In a specific implementation, the server 104 may count the number of all the human point clouds in one human point cloud set, and the number of the first target point clouds, by dividing the number of the first target point clouds by the number of all the human body point clouds, the ratio of the first target point cloud in the human body point cloud set can be obtained, a judgment result is obtained by comparing the ratio of the first target point cloud in the human body point cloud set with a preset ratio threshold, whether the falling condition exists in the space to be detected can be determined according to the judgment result, specifically, if the proportion of the first target point cloud in the human body point cloud set is above the proportion threshold value, the falling behavior of the human body corresponding to the current human body point cloud set can be determined, and if the ratio of the first target point cloud in the human body point cloud set is smaller than the ratio threshold, determining that the human body corresponding to the current human body point cloud set does not fall. For example, the server 104 may set the proportion threshold to be 80%, and 100 pieces of human body point cloud data are collected in the current human body point cloud set, where 85 pieces of first target point clouds with heights lower than 30 centimeters are collected, that is, the proportion of the first target point clouds in the human body point cloud set is 85/100-85%, and when the proportion threshold is exceeded, it may be determined that the falling behavior of the human body corresponding to the current human body point cloud set occurs.
According to the falling detection method, the spatial point cloud of the space to be detected is obtained, the human body point cloud set in the spatial point cloud is identified, and falling detection can be performed on the human body point cloud set in the space to be detected; the height of each human body point cloud in the human body point cloud set is determined according to the ground position in the space to be detected, falling detection can be carried out according to the height of the human body point cloud in the human body point cloud set, and detection accuracy is improved; if the height of the human body point cloud is lower than a preset height threshold, the human body point cloud is judged to be a first target point cloud, falling detection can be further carried out according to the first target point cloud lower than the preset height threshold, whether the proportion of the first target point cloud in the human body point cloud set exceeds a preset proportion threshold or not is judged, whether a falling condition exists in a space to be detected or not is determined according to a judgment result, the falling condition exists in the space to be detected can be judged when the human body point cloud in the human body point cloud set is generally close to the ground, and the accuracy of the detection result is improved.
In an embodiment, the step S220 may specifically include: identifying a ground range in the spatial point cloud; clustering and classifying the space point cloud according to the ground range to obtain an object point cloud set in a space to be detected; and identifying the object point cloud set to obtain a human body point cloud set in the space to be detected.
The object point cloud is point cloud data corresponding to an object, and the object point cloud set is a set formed by object point clouds in a space to be detected.
In a specific implementation, the ground range in the spatial point cloud can be identified by a spatial plane formula, which can be expressed as ax + by + cz + d being 0, by randomly selecting three points from the spatial point cloud and substituting the three points into the spatial plane formula, the plane coefficients a, b, c and d of the spatial plane formula can be determined, determining a specific expression of a space plane formula according to the plane coefficient to obtain an accurate space plane, and in order to further judge whether the space plane is the ground or not, three points can be repeatedly selected for verification through probability iteration, if the number of point clouds in the space plane exceeds a preset threshold value within a preset iteration number, the current space plane can be judged to be the ground, the current space plane formula is a ground formula, all three-dimensional data which are determined according to a certain sampling interval and accord with the ground formula can form a ground range. In the ground range, clustering and classifying all spatial point cloud data to obtain a point cloud data set corresponding to each object, and specifically, DBSCAN clustering and classifying the spatial point cloud by using an euclidean distance formula to obtain a point cloud data set corresponding to each object in each frame of depth image. For the point cloud data set corresponding to each object, a point cloud data set which may be a human body may be preliminarily screened out according to the size of the point cloud data set, and is marked as a candidate point cloud set, and then the candidate point cloud set may be identified through deep learning to detect whether the candidate point cloud set is the point cloud data set of the human body, for example, whether the candidate point cloud set is the point cloud data set of the human body may be judged through an OpenVINO deep learning technique.
In the embodiment, the detection range can be reduced to the ground range by identifying the ground range in the spatial point cloud, so that the detection complexity is reduced; clustering and classifying the spatial point clouds according to the ground range to obtain an object point cloud set in a space to be detected, so that the detection range can be further narrowed to the objects on the ground, and the detection complexity is further reduced; the human body point cloud set in the space to be detected is obtained by identifying the object point cloud set, and the falling detection accuracy can be improved by detecting the human body point cloud set.
In an embodiment, the step S220 may further include: randomly selecting a test point cloud in the space point cloud; determining a test plane corresponding to the test point cloud; marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds; judging whether the number of the second target point clouds exceeds a preset number threshold value; if not, returning to the step of randomly selecting the test point cloud in the space point cloud; and if so, obtaining the ground range in the spatial point cloud according to the test plane.
The test point cloud is randomly selected from the space point cloud data and is used for determining a space plane formula.
The test plane is a space plane calculated according to the test point cloud.
In a specific implementation, the iteration number may be set to 0 in an initial situation, three points are randomly selected from the spatial point cloud as a test point cloud, the test point cloud is substituted into an initial spatial plane formula ax + by + cz + d to 0, a joint equation set is established to obtain specific values of plane coefficients a, b, c and d in the initial spatial plane formula, a target spatial plane formula is further obtained by substituting the plane coefficients into the initial spatial plane formula, a unique spatial plane, i.e. a test plane, can be determined according to the target spatial plane formula, the distance between the spatial point cloud and the test plane is calculated, if the distance is smaller than a certain threshold, the current spatial point cloud can be determined to be the spatial point cloud in the test plane, and is marked as a second target point cloud, whether the test plane is the ground or not can be determined according to the number of the second target point cloud, specifically, if the proportion of the number of the second target point clouds to the number of all the space point clouds is lower than a certain proportion threshold value, judging that the test plane is not the ground, adding 1 to the iteration frequency, selecting three points as new test point clouds again at random, repeating the process to obtain the new second target point clouds, and if the proportion of the number of the second target point clouds to the number of all the space point clouds exceeds the proportion threshold value under the condition that the iteration frequency does not exceed a certain threshold value, judging that the current test plane is the ground, wherein the current space plane formula is a ground formula, and determining a ground range in the space point clouds according to the ground formula.
In the embodiment, test point clouds in the space point clouds are randomly selected; determining a test plane corresponding to the test point cloud; marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds; judging whether the number of the second target point clouds exceeds a preset number threshold value; if not, returning to the step of selecting the test point cloud in the space point cloud; if so, obtaining the ground range in the spatial point cloud according to the test plane, repeatedly verifying through an iteration process to finally obtain the ground range, reducing the falling detection range and reducing the falling detection complexity.
In an embodiment, the step S220 may further include: substituting point cloud data of the test point cloud into an initial space plane formula to obtain a formula coefficient of the initial space plane formula; determining a target space plane formula according to the formula coefficient; and obtaining a test plane according to a target space plane formula.
In specific implementation, three points can be randomly selected from the spatial point cloud as a test point cloud, three-dimensional data corresponding to the test point cloud is substituted into an initial spatial plane formula ax + by + cz + d as 0, a joint equation set can be established, specific numerical values of plane coefficients a, b, c and d can be obtained by solving the joint equation set, a target spatial plane formula is further obtained, and a unique spatial plane, namely the test plane, can be determined according to the target spatial plane formula.
For example, three points (x) may be randomly chosen in the spatial point cloud1,y1,z1)，(x2,y2,z2)，(x3,y3,z3) Substituting into the initial space plane formula to establishThe joint equation set can obtain a plane coefficient of a being 5, b being 8, c being-10, d being-6, further determine that a target space plane formula is 5x +8y-10z-6 being 0, sample the space to be detected according to a certain sampling interval, obtain three-dimensional sampling data of the space to be detected, wherein the three-dimensional sampling data which accord with the target space plane formula of 5x +8y-10z-6 being 0 can form a test plane.
In the embodiment, the point cloud data of the test point cloud is substituted into the initial space plane formula to obtain the formula coefficient of the initial space plane formula; determining a target space plane formula according to the formula coefficient; according to the target space plane formula, a test plane is obtained, the test plane can be obtained by selecting the test point cloud in the iteration process, and the falling detection range can be reduced and the falling detection complexity is reduced by verifying whether the test plane is the ground or not.
In an embodiment, the step S220 may further include: counting the number of the point clouds in the object point cloud set; if the number of the point clouds in the object point cloud set conforms to a preset number range, taking the object point cloud set as a candidate point cloud set; judging whether the candidate point cloud set is a human body point cloud set or not through a pre-training model; and if so, determining the candidate point cloud set as a human body point cloud set.
In a specific implementation, for the point cloud data set corresponding to each object, the number of the point cloud data in the point cloud data set may be counted, and whether the point cloud data set is likely to be a human body point cloud data set is preliminarily screened according to the number of the point cloud data, for example, the number range of the point cloud data in the human body point cloud set may be preset to [100, 200], if the number of the point cloud data in a certain object point cloud set belongs to the range, the point cloud set may be a human body point cloud set and marked as a candidate point cloud set, the point cloud in the candidate point cloud set is a candidate point cloud, and then the candidate point cloud set may be identified through a pre-training model to detect whether the point cloud set is a human body point cloud set, for example, whether the candidate point cloud set is a human body point cloud set may be determined through an OpenVINO deep learning technique, if the candidate point cloud set is a human body point cloud set, subsequent fall detection may be performed on the candidate point cloud, it is not subject to subsequent fall detection.
In the embodiment, the number of the point clouds in the object point cloud set is counted, and if the number of the point clouds in the object point cloud set meets the preset number range, the object point cloud set is used as a candidate point cloud set, so that whether the object point cloud in the space to be detected is a human point cloud or not can be preliminarily identified, the falling detection range is reduced, and the falling detection complexity is reduced; the candidate point cloud is subjected to deep learning, whether the candidate point cloud is a human body point cloud or not is judged, if yes, the human body point cloud is obtained according to the candidate point cloud, and the accuracy of falling detection can be improved through the deep learning.
In an embodiment, the step S230 may specifically include: the absolute point cloud height of the human body point cloud is obtained by calculating the distance between the human body point cloud and the ground position; and carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud.
In the concrete implementation, after the three points are repeatedly selected for verification through probability iteration and the current space plane formula is determined to be the ground formula, the position of a sampling point which accords with the ground formula in the space to be detected can be determined to be the ground position, wherein the sampling point can be three-dimensional sampling data obtained by sampling the space to be detected according to a certain sampling interval, and the position of the three-dimensional sampling data which accords with the ground formula can be the ground position. The absolute point cloud height of the human body point cloud can be obtained by calculating the distance between each human body point cloud in a human body point cloud set and the ground position, and the specific formula can be
h1＝|ax0+by0+cz0+d|，
Wherein (x)0,y0,z0) For a human point cloud data of a human point cloud set, h1Is human body point cloud data (x)0,y0,z0) Absolute point cloud height of (d). Setting normalization coefficient
The normalized coefficient is multiplied by the absolute point cloud height to obtain the point cloud height of the human body point cloud, specificallyThe formula can be
In the embodiment, the absolute point cloud height of the human body point cloud is obtained by calculating the distance between the human body point cloud and the ground position, the point cloud height of the human body point cloud is obtained by normalizing the absolute point cloud height through the preset normalization coefficient, the point cloud height of the human body point cloud can be normalized to be within a certain height range, the human body falling behavior can be conveniently identified, and the accuracy of falling detection can be improved.
In an embodiment, the step S250 may specifically include: if the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold, acquiring the duration of the condition that the ratio of the first target point cloud in the human body point cloud set exceeds the preset ratio threshold; and if the duration exceeds a preset duration threshold, determining that the falling condition exists in the space to be detected.
In specific implementation, the server can reduce erroneous judgment of fall detection by counting the duration of human body fall, the initial value of the duration can be set to be 0 second, when a preset time interval, for example, 1 second, the operation of counting the proportion of the first target point cloud in the human body point cloud set can be executed once, whether the proportion of the first target point cloud in the human body point cloud set exceeds a preset proportion threshold value or not is judged, if not, it can be determined that human body fall behavior is not detected at the current moment, human body fall detection at the next moment can be carried out, and if so, it is indicated that human body fall behavior is detected at the current moment, and the duration of human body fall can be counted. The time interval of 1 second can be used for accumulating the initial value of the duration time of 0 second, the duration time of human body falling is 1 second, whether the duration time of human body falling exceeds a preset duration threshold value or not is judged, for example, 10 seconds, if the duration time of human body falling exceeds the duration threshold value, namely, the human body continuously falls for 10 seconds, the falling condition in a space to be detected can be judged, the server can send an alarm signal to inform rescue, otherwise, if the duration time of human body falling is less than the duration threshold value, the alarm signal can not be sent, and the human body falling detection at the next moment is carried out. The method comprises the steps of judging whether the proportion of a first target point cloud in a human body point cloud set exceeds a proportion threshold or not, counting the number of all human body point clouds in one human body point cloud set and the number of the first target point clouds, obtaining the proportion of the first target point cloud in the human body point cloud set by dividing the number of the first target point cloud by the number of all human body point clouds, comparing the proportion of the first target point cloud in the human body point cloud set with a preset proportion threshold, and if the proportion of the first target point cloud in the human body point cloud set is above the proportion threshold, determining that a human body corresponding to the current human body point cloud set falls, namely, a falling condition exists in a space to be detected, otherwise, determining that the human body corresponding to the current human body point cloud set does not fall if the proportion of the first target point cloud in the human body point cloud set is smaller than the proportion threshold.
In the embodiment, if the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold, the duration that the ratio of the first target point cloud in the human body point cloud set exceeds the preset ratio threshold is obtained; if the duration exceeds the preset duration threshold, the falling condition in the space to be detected is determined, the normal bending action of the human body can be prevented from being mistakenly judged as falling, and the accuracy of falling detection can be improved.
Fig. 3 provides a schematic flow chart of another fall detection method, which specifically includes the following steps: step S310, acquiring a spatial point cloud of a space to be detected; step S320, randomly selecting a test point cloud in the space point cloud; substituting point cloud data of the test point cloud into an initial space plane formula to obtain a formula coefficient of the initial space plane formula; determining a target space plane formula according to the formula coefficient; obtaining a test plane according to a target space plane formula; marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds; judging whether the number of the second target point clouds exceeds a preset number threshold value; if not, returning to the step of randomly selecting the test point cloud in the space point cloud; if yes, obtaining a ground range in the spatial point cloud according to the test plane; step S330, clustering and classifying the spatial point cloud according to the ground range to obtain an object point cloud set in the space to be detected; step S340, counting the number of the point clouds in the object point cloud set; if the number of the point clouds in the object point cloud set conforms to a preset number range, taking the object point cloud set as a candidate point cloud set; judging whether the candidate point cloud set is a human body point cloud set or not through a pre-training model; if so, determining the candidate point cloud set as a human body point cloud set; step S350, calculating the distance between the human body point cloud and the ground position to obtain the absolute point cloud height of the human body point cloud; carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud; step S360, if the height of the human body point cloud is lower than a preset height threshold value, the human body point cloud is judged to be a first target point cloud; step S370, judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether a falling condition exists in the space to be detected according to the judgment result.
It should be understood that although the various steps in the flow charts of fig. 2-3 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2-3 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed in turn or alternately with other steps or at least some of the other steps.
In one embodiment, as shown in fig. 4, there is provided a fall detection apparatus 400 comprising: an acquisition module 401, a human body point cloud identification module 402, a height calculation module 403, a target point cloud identification module 404, and a fall detection module 405, wherein:
an obtaining module 401, configured to obtain a spatial point cloud of a space to be detected;
a human body point cloud identification module 402, configured to identify a human body point cloud set in the spatial point cloud;
a height calculation module 403, configured to determine the height of each human point cloud in the human point cloud set according to the ground position in the space to be detected;
a target point cloud identification module 404, configured to determine that the human body point cloud is a first target point cloud if the height of the human body point cloud is lower than a preset height threshold;
and the falling detection module 405 is configured to determine whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold, and determine whether a falling condition exists in the space to be detected according to a determination result.
In one embodiment, the human point cloud identification module 402 is further configured to identify a ground range in the spatial point cloud; clustering and classifying the space point cloud according to the ground range to obtain an object point cloud set in a space to be detected; and identifying the object point cloud set to obtain a human body point cloud set in the space to be detected.
In one embodiment, the human point cloud identification module 402 is further configured to randomly select a test point cloud from the spatial point clouds; determining a test plane corresponding to the test point cloud; marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds; judging whether the number of the second target point clouds exceeds a preset number threshold value; if not, returning to the step of randomly selecting the test point cloud in the space point cloud; and if so, obtaining the ground range in the spatial point cloud according to the test plane.
In an embodiment, the human point cloud identification module 402 is further configured to obtain a formula coefficient of the initial spatial plane formula by substituting the point cloud data of the test point cloud into the initial spatial plane formula; determining a target space plane formula according to the formula coefficient; and obtaining a test plane according to a target space plane formula.
In one embodiment, the human point cloud identification module 402 is further configured to count the number of point clouds in the object point cloud set; if the number of the point clouds in the object point cloud set conforms to a preset number range, taking the object point cloud set as a candidate point cloud set; judging whether the candidate point cloud set is a human body point cloud set or not through a pre-training model; and if so, determining the candidate point cloud set as a human body point cloud set.
In an embodiment, the height calculating module 403 is further configured to obtain an absolute point cloud height of the human body point cloud by calculating a distance between the human body point cloud and the ground location; and carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud.
In one embodiment, the fall detection module 405 is further configured to obtain an absolute point cloud height of the human body point cloud by calculating a distance between the human body point cloud and the ground location; and carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud.
Specific definitions of fall detection means can be found in the above definitions of fall detection methods, which are not described in detail here. The various modules in the fall detection apparatus described above may be implemented in whole or in part by software, hardware, and combinations thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used to store fall detection data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a fall detection method.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program: acquiring a space point cloud of a space to be detected; identifying a human body point cloud set in the spatial point cloud; determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected; if the height of the human body point cloud is lower than a preset height threshold value, judging that the human body point cloud is a first target point cloud; and judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the space to be detected has a falling condition according to the judgment result.
In one embodiment, the processor, when executing the computer program, further performs the steps of: identifying a ground extent in the spatial point cloud; performing clustering classification on the spatial point cloud according to the ground range to obtain an object point cloud set in the space to be detected; and identifying the object point cloud set to obtain a human body point cloud set in the space to be detected.
In one embodiment, the processor, when executing the computer program, further performs the steps of: randomly selecting a test point cloud in the space point cloud; determining a test plane corresponding to the test point cloud; marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds; judging whether the number of the second target point clouds exceeds a preset number threshold value; if not, returning to the step of randomly selecting the test point cloud in the space point cloud; and if so, obtaining the ground range in the spatial point cloud according to the test plane.
In one embodiment, the processor, when executing the computer program, further performs the steps of: substituting the point cloud data of the test point cloud into an initial space plane formula to obtain a formula coefficient of the initial space plane formula; determining a target space plane formula according to the formula coefficient; and obtaining the test plane according to the target space plane formula.
In one embodiment, the processor, when executing the computer program, further performs the steps of: counting the number of the point clouds in the object point cloud set; if the number of the point clouds in the object point cloud set conforms to a preset number range, taking the object point cloud set as a candidate point cloud set; judging whether the candidate point cloud set is a human body point cloud set or not through a pre-training model; and if so, determining the candidate point cloud set as the human body point cloud set.
In one embodiment, the processor, when executing the computer program, further performs the steps of: calculating the distance between the human body point cloud and the ground position to obtain the absolute point cloud height of the human body point cloud; and carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud.
In one embodiment, the processor, when executing the computer program, further performs the steps of: if the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold, acquiring the duration of the condition that the ratio of the first target point cloud in the human body point cloud set exceeds the preset ratio threshold; and if the duration exceeds a preset duration threshold, determining that the falling condition exists in the space to be detected.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of: acquiring a space point cloud of a space to be detected; identifying a human body point cloud set in the spatial point cloud; determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected; if the height of the human body point cloud is lower than a preset height threshold value, judging that the human body point cloud is a first target point cloud; and judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the space to be detected has a falling condition according to the judgment result.
In one embodiment, the computer program when executed by the processor further performs the steps of: identifying a ground extent in the spatial point cloud; performing clustering classification on the spatial point cloud according to the ground range to obtain an object point cloud set in the space to be detected; and identifying the object point cloud set to obtain a human body point cloud set in the space to be detected.
In one embodiment, the computer program when executed by the processor further performs the steps of: randomly selecting a test point cloud in the space point cloud; determining a test plane corresponding to the test point cloud; marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds; judging whether the number of the second target point clouds exceeds a preset number threshold value; if not, returning to the step of randomly selecting the test point cloud in the space point cloud; and if so, obtaining the ground range in the spatial point cloud according to the test plane.
In one embodiment, the computer program when executed by the processor further performs the steps of: substituting the point cloud data of the test point cloud into an initial space plane formula to obtain a formula coefficient of the initial space plane formula; determining a target space plane formula according to the formula coefficient; and obtaining the test plane according to the target space plane formula.
In one embodiment, the computer program when executed by the processor further performs the steps of: counting the number of the point clouds in the object point cloud set; if the number of the point clouds in the object point cloud set conforms to a preset number range, taking the object point cloud set as a candidate point cloud set; judging whether the candidate point cloud set is a human body point cloud set or not through a pre-training model; and if so, determining the candidate point cloud set as the human body point cloud set.
In one embodiment, the computer program when executed by the processor further performs the steps of: calculating the distance between the human body point cloud and the ground position to obtain the absolute point cloud height of the human body point cloud; and carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud.
In one embodiment, the computer program when executed by the processor further performs the steps of: if the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold, acquiring the duration of the condition that the ratio of the first target point cloud in the human body point cloud set exceeds the preset ratio threshold; and if the duration exceeds a preset duration threshold, determining that the falling condition exists in the space to be detected.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A fall detection method, characterized in that the method comprises:
acquiring a space point cloud of a space to be detected;
identifying a human body point cloud set in the spatial point cloud;
determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected;
if the height of the human body point cloud is lower than a preset height threshold value, judging that the human body point cloud is a first target point cloud;
and judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value, and determining whether the space to be detected has a falling condition according to the judgment result.
2. The method of claim 1, wherein the identifying a set of human point clouds in the spatial point cloud comprises:
identifying a ground extent in the spatial point cloud;
performing clustering classification on the spatial point cloud according to the ground range to obtain an object point cloud set in the space to be detected;
and identifying the object point cloud set to obtain a human body point cloud set in the space to be detected.
3. The method of claim 2, wherein the identifying the ground extent in the spatial point cloud comprises:
randomly selecting a test point cloud in the space point cloud;
determining a test plane corresponding to the test point cloud;
marking the space point clouds in the test plane as second target point clouds, and counting the number of the second target point clouds;
judging whether the number of the second target point clouds exceeds a preset number threshold value;
if not, returning to the step of randomly selecting the test point cloud in the space point cloud;
and if so, obtaining the ground range in the spatial point cloud according to the test plane.
4. The method of claim 3, wherein the determining the test plane corresponding to the test point cloud comprises:
substituting the point cloud data of the test point cloud into an initial space plane formula to obtain a formula coefficient of the initial space plane formula;
determining a target space plane formula according to the formula coefficient;
and obtaining the test plane according to the target space plane formula.
5. The method according to claim 2, wherein the obtaining of the cloud set of human body points in the space to be detected by identifying the cloud set of object points comprises:
counting the number of the point clouds in the object point cloud set;
if the number of the point clouds in the object point cloud set conforms to a preset number range, taking the object point cloud set as a candidate point cloud set;
judging whether the candidate point cloud set is a human body point cloud set or not through a pre-training model;
and if so, determining the candidate point cloud set as the human body point cloud set.
6. The method according to claim 4, wherein the determining the height of each human point cloud in the human point cloud set according to the ground position in the space to be detected comprises:
calculating the distance between the human body point cloud and the ground position to obtain the absolute point cloud height of the human body point cloud;
and carrying out normalization processing on the absolute point cloud height through a preset normalization coefficient to obtain the point cloud height of the human body point cloud.
7. The fall detection method according to claim 1, wherein the determining whether the fall condition exists in the space to be detected according to the determination result comprises:
if the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold, acquiring the duration of the condition that the ratio of the first target point cloud in the human body point cloud set exceeds the preset ratio threshold;
and if the duration exceeds a preset duration threshold, determining that the falling condition exists in the space to be detected.
8. A fall detection apparatus, characterized in that the apparatus comprises:
the acquisition module is used for acquiring a spatial point cloud of a space to be detected;
the human body point cloud identification module is used for identifying a human body point cloud set in the space point cloud;
the height calculation module is used for determining the height of each human body point cloud in the human body point cloud set according to the ground position in the space to be detected;
the target point cloud identification module is used for judging that the human body point cloud is a first target point cloud if the height of the human body point cloud is lower than a preset height threshold;
and the falling detection module is used for judging whether the ratio of the first target point cloud in the human body point cloud set exceeds a preset ratio threshold value or not and determining whether a falling condition exists in the space to be detected according to a judgment result.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010824761.4A CN112036269A (en) | 2020-08-17 | 2020-08-17 | Fall detection method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010824761.4A CN112036269A (en) | 2020-08-17 | 2020-08-17 | Fall detection method and device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112036269A true CN112036269A (en) | 2020-12-04 |
Family
ID=73577902
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010824761.4A Pending CN112036269A (en) | 2020-08-17 | 2020-08-17 | Fall detection method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112036269A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113947867A (en) * | 2021-09-23 | 2022-01-18 | 宁波溪棠信息科技有限公司 | Method, system, electronic device and storage medium for detecting abnormal target behaviors |
| CN114942434A (en) * | 2022-04-25 | 2022-08-26 | 西南交通大学 | Fall attitude identification method and system based on millimeter wave radar point cloud |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120025989A1 (en) * | 2010-07-30 | 2012-02-02 | General Electric Company | Method and system for detecting a fallen person using a range imaging device |
| CN109581361A (en) * | 2018-11-22 | 2019-04-05 | 九牧厨卫股份有限公司 | A kind of detection method, detection device, terminal and detection system |
| CN110909691A (en) * | 2019-11-26 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Motion detection method and device, computer readable storage medium and computer equipment |
| CN111047827A (en) * | 2019-12-03 | 2020-04-21 | 北京深测科技有限公司 | Intelligent monitoring method and system for environment-assisted life |
| CN111243230A (en) * | 2020-01-20 | 2020-06-05 | 南京邮电大学 | Human body falling detection device and method based on two depth cameras |
-
2020
- 2020-08-17 CN CN202010824761.4A patent/CN112036269A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120025989A1 (en) * | 2010-07-30 | 2012-02-02 | General Electric Company | Method and system for detecting a fallen person using a range imaging device |
| CN109581361A (en) * | 2018-11-22 | 2019-04-05 | 九牧厨卫股份有限公司 | A kind of detection method, detection device, terminal and detection system |
| CN110909691A (en) * | 2019-11-26 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Motion detection method and device, computer readable storage medium and computer equipment |
| CN111047827A (en) * | 2019-12-03 | 2020-04-21 | 北京深测科技有限公司 | Intelligent monitoring method and system for environment-assisted life |
| CN111243230A (en) * | 2020-01-20 | 2020-06-05 | 南京邮电大学 | Human body falling detection device and method based on two depth cameras |
Non-Patent Citations (1)
| Title |
|---|
| 刘松;陈小平;: "适用于家庭服务机器人的倒地人体检测方法", 计算机系统应用, no. 10 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113947867A (en) * | 2021-09-23 | 2022-01-18 | 宁波溪棠信息科技有限公司 | Method, system, electronic device and storage medium for detecting abnormal target behaviors |
| CN113947867B (en) * | 2021-09-23 | 2023-06-27 | 恒玄科技(上海)股份有限公司 | Method, system, electronic device and storage medium for detecting abnormal target behavior |
| CN114942434A (en) * | 2022-04-25 | 2022-08-26 | 西南交通大学 | Fall attitude identification method and system based on millimeter wave radar point cloud |
| CN114942434B (en) * | 2022-04-25 | 2024-02-02 | 四川八维九章科技有限公司 | Fall gesture recognition method and system based on millimeter wave Lei Dadian cloud |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109886997B (en) | Identification frame determining method and device based on target detection and terminal equipment | |
| US11308334B2 (en) | Method and apparatus for integration of detected object identifiers and semantic scene graph networks for captured visual scene behavior estimation | |
| CN103761748B (en) | Anomaly detection method and device | |
| US9792505B2 (en) | Video monitoring method, video monitoring system and computer program product | |
| CN109815787B (en) | Target identification method and device, storage medium and electronic equipment | |
| CN112926541B (en) | Sleeping post detection method and device and related equipment | |
| CN112036269A (en) | Fall detection method and device, computer equipment and storage medium | |
| AU2021203821B2 (en) | Methods, devices, apparatuses and storage media of detecting correlated objects involved in images | |
| CN112200056B (en) | Face living body detection method and device, electronic equipment and storage medium | |
| CN111079518A (en) | Fall-down abnormal behavior identification method based on scene of law enforcement and case handling area | |
| US20200211202A1 (en) | Fall detection method, fall detection apparatus and electronic device | |
| CN109344758B (en) | Face recognition method based on improved local binary pattern | |
| CN111768429A (en) | Pedestrian target tracking method in tunnel environment based on Kalman filtering and pedestrian re-identification algorithm | |
| WO2021051382A1 (en) | White balance processing method and device, and mobile platform and camera | |
| CN109005390B (en) | Method and system for establishing personnel distribution model based on signal intensity and video | |
| CN114140663A (en) | Multi-scale attention and learning network-based pest identification method and system | |
| CN111325708B (en) | Transmission line detection method and server | |
| CN111862511A (en) | Target intrusion detection device and method based on binocular stereo vision | |
| CN113947795B (en) | Mask wearing detection method, device, equipment and storage medium | |
| EP3076370B1 (en) | Method and system for selecting optimum values for parameter set for disparity calculation | |
| CN111695404B (en) | Pedestrian falling detection method and device, electronic equipment and storage medium | |
| TW201544830A (en) | Method of determining earthquake with artificial intelligence and earthquake detecting system | |
| JP7243372B2 (en) | Object tracking device and object tracking method | |
| WO2021093744A1 (en) | Method and apparatus for measuring diameter of pupil, and computer-readable storage medium | |
| CN111626193A (en) | Face recognition method, face recognition device and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |