CN118015785B - Remote monitoring nursing system and method thereof - Google Patents
Remote monitoring nursing system and method thereof Download PDFInfo
- Publication number
- CN118015785B CN118015785B CN202410404692.XA CN202410404692A CN118015785B CN 118015785 B CN118015785 B CN 118015785B CN 202410404692 A CN202410404692 A CN 202410404692A CN 118015785 B CN118015785 B CN 118015785B
- Authority
- CN
- China
- Prior art keywords
- sequence
- behavior state
- behavior
- semantic
- feature
- 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.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 154
- 238000000034 method Methods 0.000 title claims abstract description 59
- 230000000474 nursing effect Effects 0.000 title abstract description 32
- 230000006399 behavior Effects 0.000 claims abstract description 353
- 238000013135 deep learning Methods 0.000 claims abstract description 10
- 239000013598 vector Substances 0.000 claims description 48
- 239000011159 matrix material Substances 0.000 claims description 43
- 238000005457 optimization Methods 0.000 claims description 38
- 238000000605 extraction Methods 0.000 claims description 14
- 238000010586 diagram Methods 0.000 claims description 13
- 239000003623 enhancer Substances 0.000 claims description 13
- 238000007781 pre-processing Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000009826 distribution Methods 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 12
- 238000013527 convolutional neural network Methods 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 7
- 230000002787 reinforcement Effects 0.000 claims description 5
- 238000005728 strengthening Methods 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000003542 behavioural effect Effects 0.000 claims description 2
- 230000002708 enhancing effect Effects 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 5
- 238000013473 artificial intelligence Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 8
- 230000008901 benefit Effects 0.000 description 6
- 238000012549 training Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 238000003860 storage Methods 0.000 description 4
- 206010017076 Fracture Diseases 0.000 description 3
- 208000030886 Traumatic Brain injury Diseases 0.000 description 3
- 230000009471 action Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 230000036544 posture Effects 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 208000010392 Bone Fractures Diseases 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000019771 cognition Effects 0.000 description 2
- 230000009514 concussion Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 206010008111 Cerebral haemorrhage Diseases 0.000 description 1
- 206010016454 Femur fracture Diseases 0.000 description 1
- 208000001132 Osteoporosis Diseases 0.000 description 1
- 206010041541 Spinal compression fracture Diseases 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 206010048049 Wrist fracture Diseases 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000037182 bone density Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000003920 cognitive function Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000005802 health problem Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 230000033001 locomotion Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000015654 memory Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
Landscapes
- Alarm Systems (AREA)
Abstract
The application discloses a remote monitoring nursing system and a remote monitoring nursing method, which relate to the technical field of intelligent monitoring, and are characterized in that a behavior state of an old person is monitored and analyzed by utilizing an artificial intelligence technology based on deep learning, semantic time sequence coding is carried out on a behavior state monitoring video of the old person, and behavior characteristics related to a falling state in the behavior state monitoring video are dug out, so that whether the old person is in the falling state or not is intelligently judged. Therefore, the real-time monitoring and early warning of the falling event of the old can be realized, so that rescue measures can be taken in time.
Description
Technical Field
The application relates to the technical field of intelligent monitoring, in particular to a remote monitoring nursing system and a method thereof.
Background
With the development of society and the aggravation of aging population, nursing and health problems of the elderly have become a focus of social attention. Since the body functions of the aged are lowered, the balance ability, coordination ability and reaction ability are inferior to those of the young, so that a fall accident is more likely to occur.
The event of a fall of the elderly is one of the main causes of serious injury and adverse consequences. Because the old people are relatively loose, once the old people fall down, fractures such as femur fracture, spinal compression fracture, wrist fracture and the like are easy to occur, and even long-term bedridden patients can be caused, so that the life quality is seriously influenced. Moreover, the old people have high possibility of head landing when falling down, and can cause brain traumas such as concussion, cerebral hemorrhage and the like, and life can be endangered when serious. Therefore, the behavior of the old is monitored and early-warned in real time, and dangerous situations such as falling and the like are found and treated in time, so that the method has important practical significance.
However, the traditional monitoring nursing method mainly relies on manual observation and intervention, and has the problems of insufficient human resources, limited monitoring effect and the like, so that real-time monitoring and early warning are difficult to achieve. Accordingly, an optimized remote monitoring care system and method thereof are desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems.
Accordingly, according to one aspect of the present application, there is provided a remote monitoring care method comprising:
Acquiring behavior state monitoring videos of monitored elderly objects acquired by intelligent cameras installed in elderly residences;
performing data preprocessing on the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames;
transmitting the sequence of the behavior state monitoring key frames to a cloud server through a wireless network;
Performing behavior feature extraction on the sequence of the behavior state monitoring key frames at the cloud server to obtain a sequence of behavior state semantic feature graphs;
using a feature map enhancer based on a re-parameterization layer to strengthen feature information based on prior information on the sequence of the behavior state semantic feature map so as to obtain a sequence of the strengthened behavior state semantic feature map;
Determining whether to generate early warning prompt information based on time sequence features of the sequence of the enhanced behavior state semantic feature map;
Wherein determining whether to generate the early warning prompt information based on the time sequence features of the sequence of the enhanced behavior state semantic feature map comprises:
Extracting time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain a senior citizen behavior semantic time sequence coding feature vector;
the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the senile is in a falling state or not;
Responding to the classification result to generate early warning prompt information when the old people are in a falling state;
the step of extracting the time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain the senile people behavior semantic time sequence coding feature vector comprises the following steps:
optimizing each feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain a sequence of optimized enhanced behavior state semantic feature maps;
The sequence of the optimized reinforced behavior state semantic feature diagram passes through an old person behavior semantic time sequence encoder based on an LSTM model to obtain an old person behavior semantic time sequence encoding feature vector;
The optimizing the feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain the sequence of the optimized enhanced behavior state semantic feature map comprises the following steps: calculating optimization coefficients of all feature matrixes in the sequence of the enhanced behavior state semantic feature graphs; weighting and optimizing the corresponding feature matrix of the sequence of the reinforced behavior state semantic feature map by using the optimization coefficient to obtain the sequence of the optimized reinforced behavior state semantic feature map;
Specifically calculating the optimization coefficient of each feature matrix in the sequence of the enhanced behavior state semantic feature map, including: calculating the optimization coefficients of each feature matrix in the sequence of the enhanced behavior state semantic feature map according to the following optimization coefficient calculation formula; wherein, the optimization coefficient calculation formula is:
；
Wherein the method comprises the steps of Is the first feature matrix of each feature matrix in the sequence of the enhanced behavior state semantic feature graphsCharacteristic value of location,/>Probability function representing eigenvalues,/>Is the scale of each feature matrix in the sequence of the enhanced behavior state semantic feature graphs,/>Is a class probability value obtained by a classifier of the sequence of the enhanced behavior state semantic feature map, and/>Is a weight superparameter,/>Is the optimization coefficient.
In the remote monitoring and nursing method, the data preprocessing is performed on the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames, including: and performing sparse key frame sampling on the behavior state monitoring video to obtain a sequence of the behavior state monitoring key frames.
In the remote monitoring and nursing method, the cloud server performs behavior feature extraction on the sequence of the behavior state monitoring key frames to obtain a sequence of a behavior state semantic feature map, and the method comprises the following steps: and extracting state characteristics of each behavior state monitoring key frame in the sequence of the behavior state monitoring key frames by using a deep learning network model to obtain the sequence of the behavior state semantic feature map.
In the remote monitoring nursing method, the deep learning network model is an old man behavior state feature extractor based on a convolutional neural network model.
In the remote monitoring and nursing method, the step of extracting the state characteristics of each behavior state monitoring key frame in the sequence of the behavior state monitoring key frames by using a deep learning network model to obtain the sequence of the behavior state semantic characteristic map comprises the following steps: and respectively passing each behavior state monitoring key frame in the sequence of the behavior state monitoring key frames through the old man behavior state characteristic extractor based on the convolutional neural network model to obtain the sequence of the behavior state semantic characteristic diagram.
In the remote monitoring and nursing method, the feature map enhancer based on the re-parameterization layer is used for carrying out feature information enhancement based on prior information on the sequence of the behavior state semantic feature map to obtain the sequence of the enhanced behavior state semantic feature map, and the method comprises the following steps: carrying out feature map reinforcement on the behavior state semantic feature map by using the following re-parameterization formula to obtain an enhanced behavior state semantic feature map; wherein, the re-parameterization formula is:
；
Wherein, Is the mean value of the behavior state semantic feature diagram,/>Is the variance of the behavior state semantic feature map,/>Is randomly sampled from Gaussian distributionPersonal value,/>Is the/>, in the enhanced behavior state semantic feature mapAnd characteristic values.
In the above remote monitoring nursing method, the classifying result obtained by passing the senile behavior semantic time sequence coding feature vector through a classifier is used for indicating whether the senile is in a falling state or not, and the method comprises the following steps: performing full-connection coding on the behavioral semantic time sequence coding feature vector of the old people by using a full-connection layer of the classifier to obtain a full-connection coding feature vector; inputting the fully-connected coding feature vector into a Softmax classification function of the classifier to obtain probability values of the senile behavior semantic time sequence coding feature vector belonging to various classification labels, wherein the classification labels comprise that the senile is in a falling state and the senile is not in a falling state; and determining the classification label corresponding to the maximum probability value as the classification result.
According to another aspect of the present application, there is provided a remote monitoring care system comprising:
The old person behavior monitoring module is used for acquiring behavior state monitoring videos of monitored old person objects acquired by an intelligent camera installed in an old person residence;
The data preprocessing module is used for carrying out data preprocessing on the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames;
the wireless transmission module is used for transmitting the sequence of the behavior state monitoring key frames to a cloud server through a wireless network;
The behavior feature extraction module is used for extracting behavior features of the sequence of the behavior state monitoring key frames at the cloud server to obtain a sequence of a behavior state semantic feature map;
The characteristic information strengthening module is used for strengthening the characteristic information based on prior information on the sequence of the behavior state semantic characteristic map by using a characteristic map enhancer based on a re-parameterization layer so as to obtain a sequence of the strengthened behavior state semantic characteristic map;
The fall early warning module is used for determining whether early warning prompt information is generated or not based on time sequence features of the sequence of the reinforced behavior state semantic feature map;
Wherein, fall early warning module includes:
Extracting time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain a senior citizen behavior semantic time sequence coding feature vector;
the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the senile is in a falling state or not;
Responding to the classification result to generate early warning prompt information when the old people are in a falling state;
the step of extracting the time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain the senile people behavior semantic time sequence coding feature vector comprises the following steps:
optimizing each feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain a sequence of optimized enhanced behavior state semantic feature maps;
The sequence of the optimized reinforced behavior state semantic feature diagram passes through an old person behavior semantic time sequence encoder based on an LSTM model to obtain an old person behavior semantic time sequence encoding feature vector;
The optimizing the feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain the sequence of the optimized enhanced behavior state semantic feature map comprises the following steps: calculating optimization coefficients of all feature matrixes in the sequence of the enhanced behavior state semantic feature graphs; weighting and optimizing the corresponding feature matrix of the sequence of the reinforced behavior state semantic feature map by using the optimization coefficient to obtain the sequence of the optimized reinforced behavior state semantic feature map;
Specifically calculating the optimization coefficient of each feature matrix in the sequence of the enhanced behavior state semantic feature map, including: calculating the optimization coefficients of each feature matrix in the sequence of the enhanced behavior state semantic feature map according to the following optimization coefficient calculation formula; wherein, the optimization coefficient calculation formula is:
；
Wherein the method comprises the steps of Is the first feature matrix of each feature matrix in the sequence of the enhanced behavior state semantic feature graphsCharacteristic value of location,/>Probability function representing eigenvalues,/>Is the scale of each feature matrix in the sequence of the enhanced behavior state semantic feature graphs,/>Is a class probability value obtained by a classifier of the sequence of the enhanced behavior state semantic feature map, and/>Is a weight superparameter,/>Is the optimization coefficient.
Compared with the prior art, the remote monitoring nursing system and the method thereof provided by the application monitor and analyze the behavior state of the old by using the artificial intelligence technology based on deep learning, perform semantic time sequence coding on the behavior state monitoring video of the old, and dig out the behavior characteristics related to the falling state in the behavior state monitoring video so as to intelligently judge whether the old is in the falling state. Therefore, the real-time monitoring and early warning of the falling event of the old can be realized, so that rescue measures can be taken in time.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.
Fig. 1 is a flow chart of a remote monitoring care method according to an embodiment of the application.
Fig. 2 is a schematic architecture diagram of a remote monitoring care method according to an embodiment of the application.
Fig. 3 is a flowchart of determining whether to generate early warning prompt information based on time sequence features of the sequence of the enhanced behavior state semantic feature map in the remote monitoring care method according to an embodiment of the present application.
Fig. 4 is a flowchart of extracting the time sequence features of the sequence of the enhanced behavior state semantic feature map to obtain the senile people behavior semantic time sequence coding feature vector in the remote monitoring nursing method according to the embodiment of the application.
Fig. 5 is a flowchart of the method for remote monitoring and nursing according to an embodiment of the present application, wherein the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result.
Fig. 6 is a block diagram of a remote monitoring care system according to an embodiment of the present application.
Detailed Description
For an understanding of embodiments of the present invention, specific embodiments of the invention will be described in more detail below with reference to the drawings. While specific embodiments of the invention are shown in the drawings, it should be understood that the invention may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
Fig. 1 is a flow chart of a remote monitoring care method according to an embodiment of the application. Fig. 2 is a schematic architecture diagram of a remote monitoring care method according to an embodiment of the application. As shown in fig. 1 and 2, the remote monitoring nursing method according to the embodiment of the application includes the steps of: s110, acquiring a behavior state monitoring video of a monitored old person object acquired by an intelligent camera installed in an old person residence; s120, carrying out data preprocessing on the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames; s130, transmitting the sequence of the behavior state monitoring key frames to a cloud server through a wireless network; s140, at the cloud server, performing behavior feature extraction on the sequence of the behavior state monitoring key frames to obtain a sequence of a behavior state semantic feature map; s150, carrying out feature information reinforcement based on prior information on the sequence of the behavior state semantic feature map by using a feature map enhancer based on a re-parameterization layer so as to obtain a sequence of the reinforced behavior state semantic feature map; s160, determining whether to generate early warning prompt information based on the time sequence features of the sequence of the enhanced behavior state semantic feature map.
As described above in the background art, a fall is a common accident that is extremely harmful to the health of the elderly. The body functions of the old are gradually reduced, so that the balance capacity and the muscle strength are weakened, and the old is easier to fall down. And the old may have serious consequences such as fracture, brain trauma and the like once falling down. In particular, the old has a reduced bone density and an increased risk of osteoporosis, and therefore a higher probability of fracture when falling. Also, if the elderly's head is impacted while falling, it may cause concussion or other brain trauma, which may have a long-term impact on the cognitive function and daily life ability of the elderly. Therefore, the behavior of the old is monitored and early-warned in real time, and dangerous situations such as falling and the like are found and treated in time, so that the method has important practical significance. However, the traditional elderly monitoring nursing method mainly relies on manual observation and intervention of nursing staff or family members, and has the problems of insufficient human resources, limited monitoring effect and the like, so that real-time monitoring and early warning are difficult to achieve.
Aiming at the technical problems, the technical concept of the application is to monitor and analyze the behavior state of the old people by using an artificial intelligence technology based on deep learning, perform semantic time sequence coding on a behavior state monitoring video of the old people, and dig out behavior characteristics related to the falling state in the behavior state monitoring video so as to intelligently judge whether the old people are in the falling state. Therefore, the real-time monitoring and early warning of the falling event of the old can be realized, so that rescue measures can be taken in time.
In the remote monitoring and nursing method, in step S110, a behavior state monitoring video of a monitored elderly person object collected by an intelligent camera installed in an elderly person residence is obtained. It should be understood that the intelligent camera can realize the remote monitoring to the old person's action, has important meaning to unable family members or the nursing staff who takes care of the old person in real time. Family members or nursing staff can know the condition of the elderly through videos, and provide necessary assistance and support for the elderly in time. Moreover, the behavior state monitoring video acquired by the intelligent camera provides information such as the posture, the action and the walking mode of the old person, is a data base of behavior state monitoring analysis, further carries out real-time analysis and processing on the video through a computer vision technology, can mine behavior features related to the falling state, such as sudden inclination, unbalance and the like, and further intelligently judges whether the old person is in the falling state or not so as to monitor and judge the behavior state of the old person.
In the remote monitoring and nursing method, in step S120, the behavior state monitoring video is subjected to data preprocessing to obtain a sequence of behavior state monitoring key frames. In a specific example of the present application, the data preprocessing of the behavior state monitoring video to obtain the sequence of behavior state monitoring key frames is performed by sampling sparse key frames of the behavior state monitoring video to obtain the sequence of behavior state monitoring key frames. It should be appreciated that it is contemplated that the behavior state monitoring video is a continuous video stream that includes a large number of frame images. However, since the time interval between adjacent image frames in the behavior state monitoring video is short, the difference between the adjacent image frames is small, a large amount of redundant information exists, if the whole video stream is directly processed, the data volume is huge, the calculation and storage cost is increased, and the extraction effect of the behavior characteristics is affected. Therefore, in order to reduce the data volume and improve the processing efficiency, the behavior state monitoring video is further subjected to sparse key frame sampling to obtain a sequence of behavior state monitoring key frames. That is, a part of the key frames in the behavior state monitoring video are selected for processing in a sparse key frame sampling mode, so that the processed data size is reduced, and the behavior state information in the behavior state monitoring video is extracted, so that the calculation and storage requirements of a system are reduced while the key behavior information is maintained, and the instantaneity and the efficiency of the system are improved.
In the remote monitoring and nursing method, in step S130, the sequence of the behavior state monitoring key frames is transmitted to a cloud server through a wireless network. It should be appreciated that the cloud server may provide reliable data storage and backup functions. After the sequence of the behavior state monitoring key frames is transmitted to the cloud server, the data can be stored and backed up for a long time, and the data is prevented from being lost or damaged. Moreover, cross-region access and management can be realized through the cloud server. Thus, family members, nursing staff or medical staff can be connected to the cloud server through the Internet, and behavior states of the old can be accessed and monitored anytime and anywhere. Meanwhile, the cloud server has strong computing power and storage resources, can provide computing and storage resources for behavior feature analysis processing of the sequence of the behavior state monitoring key frames, and provides more convenience and advantages for remote monitoring and nursing of the old.
In the remote monitoring and nursing method, in step S140, at the cloud server, a behavior feature extraction is performed on the sequence of the behavior state monitoring key frames to obtain a sequence of a behavior state semantic feature map. In a specific example of the present application, the sequence of behavior state monitoring key frames is encoded in such a way that the behavior feature extraction is performed on the sequence of behavior state monitoring key frames to obtain the sequence of behavior state semantic feature graphs, where each behavior state monitoring key frame in the sequence of behavior state monitoring key frames passes through the old man behavior state feature extractor based on the convolutional neural network model, so as to obtain the sequence of behavior state semantic feature graphs. Those of ordinary skill in the art will appreciate that convolutional neural networks have good characterizations in the field of image processing, and can automatically learn and extract important features in an input image through convolutional and pooling layers. In the technical scheme of the application, the old man behavior state feature extractor based on the convolutional neural network model is used for respectively carrying out image semantic feature mining on each behavior state monitoring key frame in the sequence of the behavior state monitoring key frames, so that the behavior state semantic features of the old man can be extracted, and the information such as actions, postures and movement modes of the old man can be captured, thereby providing a valuable information basis for subsequent behavior state classification and identification.
In the remote monitoring and caring method, in step S150, feature information reinforcement based on prior information is performed on the sequence of behavior state semantic feature graphs by using a feature graph enhancer based on a re-parameterization layer to obtain a sequence of reinforced behavior state semantic feature graphs. It should be appreciated that the re-parameterization layer is a technique commonly used in neural networks to generate learnable parameters from potential space, which are considered a priori information representations, and thus parameter updates via back propagation, to achieve re-parameterization of the input data. In the technical scheme of the application, the sequence of the behavior state semantic feature map is input into a feature map enhancer based on a re-parameterization layer, the re-parameterization layer is used for converting the mean value and the variance of the behavior state semantic feature map into two learnable parameters, and parameter updating is realized by randomly sampling from Gaussian distribution, so that the behavior state semantic feature map is reconstructed. Here, the mean and variance of the behavior state semantic feature graphs are considered a priori information representations. By introducing prior information, the feature map enhancer can utilize the prior information to adjust the feature distribution of the feature map, so that the feature values in the semantic feature map of each enhanced behavior state have randomness, and the effect of enhancing the data in the semantic feature space is skillfully achieved. It is worth mentioning that, because of the randomness of Gaussian distribution sampling, the feature map obtained during each reconstruction has randomness, can adapt to changes such as different illumination, postures, backgrounds and the like, further can better represent key information and dynamic changes of behavior states, and improves accuracy and reliability of behavior state monitoring.
In a specific example of the present application, the step S150, using a feature map enhancer based on a re-parameterization layer, performs feature information enhancement based on a priori information on the sequence of behavior state semantic feature maps to obtain a sequence of enhanced behavior state semantic feature maps, includes: carrying out feature map reinforcement on the behavior state semantic feature map by using the following re-parameterization formula to obtain an enhanced behavior state semantic feature map; wherein, the re-parameterization formula is:
；
Wherein, Is the mean value of the behavior state semantic feature diagram,/>Is the variance of the behavior state semantic feature map,/>Is randomly sampled from Gaussian distributionPersonal value,/>Is the/>, in the enhanced behavior state semantic feature mapAnd characteristic values.
In the remote monitoring and nursing method, step S160 determines whether to generate the early warning prompt information based on the time sequence features of the sequence of the enhanced behavior state semantic feature map. Specifically, fig. 3 is a flowchart of determining whether to generate early warning prompt information based on time sequence features of the sequence of the enhanced behavior state semantic feature graphs in the remote monitoring care method according to an embodiment of the present application. As shown in fig. 3, the step S160 includes: s161, extracting time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain senile people behavior semantic time sequence coding feature vectors; s162, enabling the senile behavior semantic time sequence coding feature vector to pass through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the senile is in a falling state or not; and S163, generating early warning prompt information in response to the classification result that the old people are in a falling state.
Specifically, in step S161, the sequential features of the sequence of the enhanced behavior state semantic feature map are extracted to obtain the senile people behavior semantic sequential coding feature vector. It should be understood that the behavior state of the elderly is a dynamically changing process, and the convolution operation and the feature map enhancement operation described above are behavior state analysis processes performed independently on the behavior state monitoring key frames at each time point. That is, the sequence of the enhanced behavior state semantic feature map has time-sequential dynamic change information in the time dimension. Therefore, in order to capture the time sequence evolution information of the behavior state of the aged people, the sequence of the semantic feature map of the enhanced behavior state is further subjected to up-down Wen Shixu association coding so as to capture the time sequence information and the change trend of the behavior state of the aged people.
Fig. 4 is a flowchart of extracting the time sequence features of the sequence of the enhanced behavior state semantic feature map to obtain the senile people behavior semantic time sequence coding feature vector in the remote monitoring nursing method according to the embodiment of the application. As shown in fig. 4, the step S161 includes: s1611, optimizing each feature matrix in the sequence of the reinforced behavior state semantic feature map to obtain a sequence of the optimized reinforced behavior state semantic feature map; s1612, passing the sequence of the optimized reinforced behavior state semantic feature map through an old person behavior semantic time sequence encoder based on an LSTM model to obtain the old person behavior semantic time sequence encoding feature vector.
Specifically, in the step S1611, each feature matrix in the sequence of the enhanced behavior state semantic feature graphs is optimized to obtain a sequence of optimized enhanced behavior state semantic feature graphs. It should be understood that, in the technical solution of the present application, each enhanced behavior state semantic feature map of the sequence of enhanced behavior state semantic feature maps expresses an image semantic feature of a key frame of the behavior state monitoring video, where, considering that when the behavior state monitoring video is sampled by a sparse key frame, the obtained key frame of the behavior state monitoring video expresses sparsity for a source image semantic time sequence distribution of the behavior state monitoring video, when the feature map enhancer based on the re-parameterization layer strengthens feature information of the sequence of behavior state semantic feature maps based on image semantic prior information in units of feature matrices, a significant channel time sequence distribution sparsity of the sequence of enhanced behavior state semantic feature maps results, which leads to a poor feature extraction effect when the behavior state semantic time sequence encoder based on an LSTM model performs time sequence encoding based on a time sequence short-long-range dual context, affects a probabilistic regression effect based on a feature value overall numerical distribution of the obtained behavior time sequence encoding feature vector of the elderly, that is, affects a training result classification of the training speed of the behavior state semantic feature vector regression of the elderly person, and obtains an accurate result classification. Therefore, the application optimizes each feature matrix in the sequence of the enhanced behavior state semantic feature map.
Specifically, the step S1611 includes: calculating optimization coefficients of all feature matrixes in the sequence of the enhanced behavior state semantic feature graphs; and carrying out weighted optimization on the corresponding feature matrix of the sequence of the enhanced behavior state semantic feature graphs by using the optimization coefficient to obtain the sequence of the optimized enhanced behavior state semantic feature graphs.
Specifically, calculating the optimization coefficient of each feature matrix in the sequence of the enhanced behavior state semantic feature map includes: calculating the optimization coefficients of each feature matrix in the sequence of the enhanced behavior state semantic feature map according to the following optimization coefficient calculation formula; wherein, the optimization coefficient calculation formula is:
；
Wherein the method comprises the steps of Is the first feature matrix of each feature matrix in the sequence of the enhanced behavior state semantic feature graphsCharacteristic value of location,/>Probability function representing eigenvalues,/>Is the scale of each feature matrix in the sequence of the enhanced behavior state semantic feature graphs,/>Is a class probability value obtained by a classifier of the sequence of the enhanced behavior state semantic feature map, and/>Is a weight superparameter,/>Is the optimization coefficient.
That is, for each feature matrix corresponding to the sequence of the enhanced behavior state semantic feature graphs, a class probability reasoning logic association of scene saturation is accepted through probability distribution foreground constraint and relative probability mapping response assumption, so that scene concept ontology cognition is given to the feature set of each feature matrix in the sequence of the enhanced behavior state semantic feature graphs, that is, integral distribution and scene-based class probability logic reasoning under the classification process are internally aligned, so that the understanding capability of the feature matrix scene distribution in the sequence of the enhanced behavior state semantic feature graphs on the class cognition is improved. Thus, the optimization coefficient is used againAnd carrying out weighted optimization on the corresponding feature matrix in the sequence of the enhanced behavior state semantic feature map, so that the quasi-probability regression effect of the optimized senile behavior semantic time sequence coding feature vector can be improved, namely, the training speed of the senile behavior semantic time sequence coding feature vector for classifying training through a classifier and the accuracy of the obtained classification result are improved.
Specifically, in step S1612, the sequence of the optimized enhanced behavior state semantic feature map is passed through an senile behavior semantic time sequence encoder based on an LSTM model to obtain the senile behavior semantic time sequence encoding feature vector. Those of ordinary skill in the art will appreciate that LSTM (Long Short-Term Memory) is a variant of recurrent neural network (Recurrent Neural Network, RNN) with strong sequence data processing capability that is capable of efficiently capturing and storing Long-Term dependencies in sequence data through Memory units and gating mechanisms. In the technical scheme of the application, the senile behavior semantic time sequence encoder uses an LSTM model as a basis, takes the sequence of the optimized reinforced behavior state semantic feature map as input, and learns and extracts time sequence evolution information of the senile behavior state, including the sequence, duration, frequency and the like of behaviors by gradually processing the optimized reinforced behavior state semantic feature map of each time step. Specifically, in each time step, the LSTM model updates the current hidden state according to the current input feature map and the hidden state of the previous time step, and takes the hidden state of the last time step as output, so as to obtain the semantic time sequence coding feature vector of the behavior of the elderly people, which includes the time sequence evolution trend of the behavior state of the elderly people and key features.
Specifically, in step S162, the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result, where the classification result is used to indicate whether the senile is in a falling state. It should be understood that in the technical scheme of the application, based on the supervised learning method, the labeled tumbling behavior and non-tumbling behavior sample data are used for training and optimizing the classifier, so that the classifier learns the mapping relation between different behavior state features and the predefined classification labels, and the accurate recognition of the tumbling behavior is realized. Furthermore, the senile behavior semantic time sequence coding feature vector is input into a classifier which is trained, the classifier can learn the behavior feature representation in the senile behavior semantic time sequence coding feature vector, map the senile behavior semantic time sequence coding feature vector into corresponding behavior type labels according to the mapping relation learned in the training process, and output a classification result, namely, the senile is in a falling state or the senile is not in the falling state.
Fig. 5 is a flowchart of the method for remote monitoring and nursing according to an embodiment of the present application, wherein the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result. As shown in fig. 5, the step S162 includes: s1621, performing full-connection coding on the senile behavior semantic time sequence coding feature vector by using a full-connection layer of the classifier to obtain a full-connection coding feature vector; s1622, inputting the fully-connected coding feature vector into a Softmax classification function of the classifier to obtain probability values of the senile behavior semantic time sequence coding feature vector belonging to various classification labels, wherein the classification labels comprise that the senile is in a falling state and the senile is not in a falling state; s1623, determining the classification label corresponding to the maximum probability value as the classification result.
Specifically, in step S163, early warning prompt information is generated in response to the classification result that the elderly person is in a fall state. It should be appreciated that if the classification result indicates that the elderly is in a fall, this means that the elderly may have a fall event and require timely assistance and assistance. Therefore, the system can respond to the classification result to generate early warning prompt information, such as sending a short message to nursing staff, family members, medical institutions or other related staff, so as to ensure that the old can be timely assisted and focused. Therefore, the real-time detection and early warning of the falling event of the old people can be realized, the related personnel can be timely reminded to pay attention to the safety condition of the old people, and timely rescue measures are provided for the old people.
In summary, the remote monitoring nursing method according to the embodiment of the application is explained, which monitors and analyzes the behavior state of the elderly by using an artificial intelligence technology based on deep learning, performs semantic time sequence coding on a behavior state monitoring video of the elderly, and digs out behavior features related to the falling state in the behavior state monitoring video so as to intelligently judge whether the elderly is in the falling state. Therefore, the real-time monitoring and early warning of the falling event of the old can be realized, so that rescue measures can be taken in time.
Fig. 6 is a block diagram of a remote monitoring care system according to an embodiment of the present application. As shown in fig. 6, a remote monitoring care system 100 according to an embodiment of the present application includes: the old people behavior monitoring module 110 is used for acquiring behavior state monitoring videos of monitored old people objects acquired by an intelligent camera installed in an old people residence; the data preprocessing module 120 is configured to perform data preprocessing on the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames; the wireless transmission module 130 is configured to transmit the sequence of the behavior state monitoring key frames to a cloud server through a wireless network; the behavior feature extraction module 140 is configured to perform behavior feature extraction on the sequence of the behavior state monitoring key frames at the cloud server to obtain a sequence of a behavior state semantic feature map; the feature information strengthening module 150 is configured to strengthen feature information based on prior information on the sequence of the behavior state semantic feature graphs by using a feature graph enhancer based on a re-parameterization layer to obtain a sequence of strengthened behavior state semantic feature graphs; and the fall early warning module 160 is configured to determine whether to generate early warning prompt information based on the time sequence features of the sequence of the reinforced behavior state semantic feature map.
Here, it will be understood by those skilled in the art that the specific operations of the respective modules in the above-described remote monitoring care system have been described in detail in the above description of the remote monitoring care method with reference to fig. 1 to 5, and thus, repetitive descriptions thereof will be omitted.
The basic principles of the present invention have been described above in connection with specific embodiments, but it should be noted that the advantages, benefits, effects, etc. mentioned in the present invention are merely examples and not intended to be limiting, and these advantages, benefits, effects, etc. are not to be construed as necessarily possessed by the various embodiments of the invention. Furthermore, the particular details of the above-described embodiments are for purposes of illustration and understanding only, and are not intended to limit the invention to the particular details described above, but are not necessarily employed.
In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods may be implemented in other ways. For example, the system embodiments described above are merely illustrative, and for example, the module division is merely a logical function division, and other manners of division may be implemented in practice. The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
Furthermore, it is evident that the word "comprising" does not exclude other elements or steps, and that the singular does not exclude a plurality. The units recited in the system claims may also be implemented by means of software or hardware.
Finally, it should be noted that the foregoing description has been presented for the purposes of illustration and description. Furthermore, the foregoing embodiments are merely for illustrating the technical scheme of the present invention, and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the technical scheme of the present invention.
Claims (8)
1. A method of remote monitoring care comprising:
Acquiring behavior state monitoring videos of monitored elderly objects acquired by intelligent cameras installed in elderly residences;
performing data preprocessing on the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames;
transmitting the sequence of the behavior state monitoring key frames to a cloud server through a wireless network;
Performing behavior feature extraction on the sequence of the behavior state monitoring key frames at the cloud server to obtain a sequence of behavior state semantic feature graphs;
using a feature map enhancer based on a re-parameterization layer to strengthen feature information based on prior information on the sequence of the behavior state semantic feature map so as to obtain a sequence of the strengthened behavior state semantic feature map;
Determining whether to generate early warning prompt information based on time sequence features of the sequence of the enhanced behavior state semantic feature map;
Wherein determining whether to generate the early warning prompt information based on the time sequence features of the sequence of the enhanced behavior state semantic feature map comprises:
Extracting time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain a senior citizen behavior semantic time sequence coding feature vector;
the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the senile is in a falling state or not;
Responding to the classification result to generate early warning prompt information when the old people are in a falling state;
the step of extracting the time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain the senile people behavior semantic time sequence coding feature vector comprises the following steps:
optimizing each feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain a sequence of optimized enhanced behavior state semantic feature maps;
The sequence of the optimized reinforced behavior state semantic feature diagram passes through an old person behavior semantic time sequence encoder based on an LSTM model to obtain an old person behavior semantic time sequence encoding feature vector;
The optimizing the feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain the sequence of the optimized enhanced behavior state semantic feature map comprises the following steps: calculating optimization coefficients of all feature matrixes in the sequence of the enhanced behavior state semantic feature graphs; weighting and optimizing the corresponding feature matrix of the sequence of the reinforced behavior state semantic feature map by using the optimization coefficient to obtain the sequence of the optimized reinforced behavior state semantic feature map;
Specifically calculating the optimization coefficient of each feature matrix in the sequence of the enhanced behavior state semantic feature map, including: calculating the optimization coefficients of each feature matrix in the sequence of the enhanced behavior state semantic feature map according to the following optimization coefficient calculation formula; wherein, the optimization coefficient calculation formula is:
；
Wherein the method comprises the steps of Is the/>, of each feature matrix in the sequence of the enhanced behavior state semantic feature graphsCharacteristic value of location,/>Probability function representing eigenvalues,/>Is the scale of each feature matrix in the sequence of the enhanced behavior state semantic feature graphs,/>Is a class probability value obtained by a classifier of the sequence of the enhanced behavior state semantic feature map, and/>Is a weight superparameter,/>Is the optimization coefficient.
2. The remote monitoring care method of claim 1, wherein the data preprocessing of the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames comprises:
and performing sparse key frame sampling on the behavior state monitoring video to obtain a sequence of the behavior state monitoring key frames.
3. The remote monitoring care method according to claim 2, wherein the performing, at the cloud server, behavior feature extraction on the sequence of behavior state monitoring key frames to obtain a sequence of behavior state semantic feature graphs includes:
and extracting state characteristics of each behavior state monitoring key frame in the sequence of the behavior state monitoring key frames by using a deep learning network model to obtain the sequence of the behavior state semantic feature map.
4. The remote monitoring care method according to claim 3, wherein the deep learning network model is an old man behavior state feature extractor based on a convolutional neural network model.
5. The remote monitoring care method as claimed in claim 4, wherein performing state feature extraction on each behavior state monitoring key frame in the sequence of behavior state monitoring key frames by using a deep learning network model to obtain the sequence of behavior state semantic feature graphs comprises:
and respectively passing each behavior state monitoring key frame in the sequence of the behavior state monitoring key frames through the old man behavior state characteristic extractor based on the convolutional neural network model to obtain the sequence of the behavior state semantic characteristic diagram.
6. The remote monitoring care method according to claim 5, wherein the step of enhancing the sequence of behavior state semantic feature graphs with the prior information-based feature information using a re-parameterized layer-based feature graph enhancer to obtain the sequence of enhanced behavior state semantic feature graphs comprises:
carrying out feature map reinforcement on the behavior state semantic feature map by using the following re-parameterization formula to obtain an enhanced behavior state semantic feature map; wherein, the re-parameterization formula is:
；
Wherein, Is the mean value of the behavior state semantic feature diagram,/>Is the variance of the behavior state semantic feature map,/>Is randomly sampled from Gaussian distributionPersonal value,/>Is the/>, in the enhanced behavior state semantic feature mapAnd characteristic values.
7. The remote monitoring care method according to claim 6, wherein the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result, and the classification result is used for indicating whether the senile is in a falling state or not, and the method comprises the following steps:
Performing full-connection coding on the behavioral semantic time sequence coding feature vector of the old people by using a full-connection layer of the classifier to obtain a full-connection coding feature vector;
Inputting the fully-connected coding feature vector into a Softmax classification function of the classifier to obtain probability values of the senile behavior semantic time sequence coding feature vector belonging to various classification labels, wherein the classification labels comprise that the senile is in a falling state and the senile is not in a falling state;
and determining the classification label corresponding to the maximum probability value as the classification result.
8. A remote monitoring care system, comprising:
The old person behavior monitoring module is used for acquiring behavior state monitoring videos of monitored old person objects acquired by an intelligent camera installed in an old person residence;
The data preprocessing module is used for carrying out data preprocessing on the behavior state monitoring video to obtain a sequence of behavior state monitoring key frames;
the wireless transmission module is used for transmitting the sequence of the behavior state monitoring key frames to a cloud server through a wireless network;
The behavior feature extraction module is used for extracting behavior features of the sequence of the behavior state monitoring key frames at the cloud server to obtain a sequence of a behavior state semantic feature map;
The characteristic information strengthening module is used for strengthening the characteristic information based on prior information on the sequence of the behavior state semantic characteristic map by using a characteristic map enhancer based on a re-parameterization layer so as to obtain a sequence of the strengthened behavior state semantic characteristic map;
The fall early warning module is used for determining whether early warning prompt information is generated or not based on time sequence features of the sequence of the reinforced behavior state semantic feature map;
Wherein, fall early warning module includes:
Extracting time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain a senior citizen behavior semantic time sequence coding feature vector;
the senile behavior semantic time sequence coding feature vector is passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the senile is in a falling state or not;
Responding to the classification result to generate early warning prompt information when the old people are in a falling state;
the step of extracting the time sequence features of the sequence of the reinforced behavior state semantic feature map to obtain the senile people behavior semantic time sequence coding feature vector comprises the following steps:
optimizing each feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain a sequence of optimized enhanced behavior state semantic feature maps;
The sequence of the optimized reinforced behavior state semantic feature diagram passes through an old person behavior semantic time sequence encoder based on an LSTM model to obtain an old person behavior semantic time sequence encoding feature vector;
The optimizing the feature matrix in the sequence of the enhanced behavior state semantic feature map to obtain the sequence of the optimized enhanced behavior state semantic feature map comprises the following steps: calculating optimization coefficients of all feature matrixes in the sequence of the enhanced behavior state semantic feature graphs; weighting and optimizing the corresponding feature matrix of the sequence of the reinforced behavior state semantic feature map by using the optimization coefficient to obtain the sequence of the optimized reinforced behavior state semantic feature map;
Specifically calculating the optimization coefficient of each feature matrix in the sequence of the enhanced behavior state semantic feature map, including: calculating the optimization coefficients of each feature matrix in the sequence of the enhanced behavior state semantic feature map according to the following optimization coefficient calculation formula; wherein, the optimization coefficient calculation formula is:
；
Wherein the method comprises the steps of Is the/>, of each feature matrix in the sequence of the enhanced behavior state semantic feature graphsCharacteristic value of location,/>Probability function representing eigenvalues,/>Is the scale of each feature matrix in the sequence of the enhanced behavior state semantic feature graphs,/>Is a class probability value obtained by a classifier of the sequence of the enhanced behavior state semantic feature map, and/>Is a weight superparameter,/>Is the optimization coefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410404692.XA CN118015785B (en) | 2024-04-07 | 2024-04-07 | Remote monitoring nursing system and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410404692.XA CN118015785B (en) | 2024-04-07 | 2024-04-07 | Remote monitoring nursing system and method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118015785A CN118015785A (en) | 2024-05-10 |
| CN118015785B true CN118015785B (en) | 2024-06-07 |
Family
ID=90948739
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410404692.XA Active CN118015785B (en) | 2024-04-07 | 2024-04-07 | Remote monitoring nursing system and method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118015785B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2015057630A (en) * | 2013-08-13 | 2015-03-26 | 日本電信電話株式会社 | Acoustic event identification model learning device, acoustic event detection device, acoustic event identification model learning method, acoustic event detection method, and program |
| CN108931770A (en) * | 2018-05-30 | 2018-12-04 | 西安电子科技大学 | ISAR imaging method based on multidimensional beta process linear regression |
| WO2021212883A1 (en) * | 2020-04-20 | 2021-10-28 | 电子科技大学 | Fall detection method based on intelligent mobile terminal |
| CN117315886A (en) * | 2023-09-07 | 2023-12-29 | 安徽建筑大学 | UWB radar-based method and device for detecting impending falling of personnel |
| CN117409538A (en) * | 2023-12-13 | 2024-01-16 | 吉林大学 | Wireless fall-prevention alarm system and method for nursing |
| CN117676099A (en) * | 2024-02-01 | 2024-03-08 | 深圳市丛文安全电子有限公司 | Security early warning method and system based on Internet of things |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117924956A (en) * | 2024-03-20 | 2024-04-26 | 内蒙古安特威盾防水科技有限公司 | Preparation process of weather-proof modified asphalt waterproof coiled material added with rare earth elements |
| CN117934463A (en) * | 2024-03-21 | 2024-04-26 | 山东农业工程学院 | Beef cattle carcass quality grading method based on optical test |
-
2024
- 2024-04-07 CN CN202410404692.XA patent/CN118015785B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2015057630A (en) * | 2013-08-13 | 2015-03-26 | 日本電信電話株式会社 | Acoustic event identification model learning device, acoustic event detection device, acoustic event identification model learning method, acoustic event detection method, and program |
| CN108931770A (en) * | 2018-05-30 | 2018-12-04 | 西安电子科技大学 | ISAR imaging method based on multidimensional beta process linear regression |
| WO2021212883A1 (en) * | 2020-04-20 | 2021-10-28 | 电子科技大学 | Fall detection method based on intelligent mobile terminal |
| CN117315886A (en) * | 2023-09-07 | 2023-12-29 | 安徽建筑大学 | UWB radar-based method and device for detecting impending falling of personnel |
| CN117409538A (en) * | 2023-12-13 | 2024-01-16 | 吉林大学 | Wireless fall-prevention alarm system and method for nursing |
| CN117676099A (en) * | 2024-02-01 | 2024-03-08 | 深圳市丛文安全电子有限公司 | Security early warning method and system based on Internet of things |
Non-Patent Citations (3)
| Title |
|---|
| 《 Fall Risk Assessment Using Pressure Insole Sensors and Convolutional Neural Networks》;Brome, Reem Nasreddine, Jad;《2022 IEEE-EMBS Conference on Biomedical Engineering and Sciences》;20221231;全文 * |
| 基于北斗RDSS的核辐射监测应急通讯方法;王廷银;林明贵;陈达;吴允平;;计算机系统应用;20191215(第12期);全文 * |
| 基于独立循环神经网络的跌倒检测方法;王晶晶;黄勇;陈宝欣;赵运勇;;实验室研究与探索;20200725(第07期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118015785A (en) | 2024-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111079655B (en) | Method for recognizing human body behaviors in video based on fusion neural network | |
| CN112016435A (en) | Abnormal behavior detection method and device, electronic equipment and storage medium | |
| Xu et al. | Intelligent emotion detection method based on deep learning in medical and health data | |
| CN107967941A (en) | A kind of unmanned plane health monitoring method and system based on intelligent vision reconstruct | |
| Patnam et al. | Deep learning based recognition of meltdown in autistic kids | |
| CN113096819A (en) | Epidemic situation prevention, control, screening and early warning system based on neural convolutional network | |
| CN116503785A (en) | Natatorium supervision system and natatorium supervision method | |
| Zhu et al. | Falling motion detection algorithm based on deep learning | |
| CN116269351A (en) | Cerebral apoplexy patient rehabilitation effect evaluation system based on movement posture change | |
| Karim et al. | Predicting autism spectrum disorder (ASD) meltdown using fuzzy semi-supervised learning with NNRW | |
| Lin et al. | Adaptive multi-modal fusion framework for activity monitoring of people with mobility disability | |
| Meena et al. | Seq2Dense U-Net: Analysing Sequential Inertial Sensor data for Human Activity Recognition using Dense Segmentation Model | |
| CN116977330B (en) | Atrial fibrillation auxiliary analysis method based on pulse neural network and context awareness | |
| CN118015785B (en) | Remote monitoring nursing system and method thereof | |
| CN116758355A (en) | Image classification method and device, electronic equipment and storage medium | |
| Sri et al. | A relative analysis of machine learning based approaches to detect human pain intensity using facial expressions | |
| En et al. | Computer-aided knee osteoarthritis classification system using artificial neural netwrok (ANN) | |
| CN113688790A (en) | Human body action early warning method and system based on image recognition | |
| CN111554398A (en) | Remote vital sign evaluation method and system based on 5G | |
| Qiu et al. | Driving anomaly detection using conditional generative adversarial network | |
| Anderson et al. | Extension of a soft-computing framework for activity analysis from linguistic summarizations of video | |
| CN117158904B (en) | Old people cognitive disorder detection system and method based on behavior analysis | |
| CN117643461B (en) | Heart rate intelligent monitoring system and method based on artificial intelligence | |
| Wang et al. | Chaos in Motion: Unveiling Robustness in Remote Heart Rate Measurement through Brain-Inspired Skin Tracking | |
| CN117727072B (en) | Patient care safety informatization management system and method |
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 | ||
| GR01 | Patent grant |