CN114114223A - Millimeter wave sitting posture detection intelligent desk lamp based on convolutional neural network - Google Patents
Millimeter wave sitting posture detection intelligent desk lamp based on convolutional neural network Download PDFInfo
- Publication number
- CN114114223A CN114114223A CN202111448426.XA CN202111448426A CN114114223A CN 114114223 A CN114114223 A CN 114114223A CN 202111448426 A CN202111448426 A CN 202111448426A CN 114114223 A CN114114223 A CN 114114223A
- Authority
- CN
- China
- Prior art keywords
- target
- sitting posture
- distance
- information
- neural network
- 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 26
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 23
- 230000036544 posture Effects 0.000 claims abstract description 54
- 238000012549 training Methods 0.000 claims description 15
- 238000012360 testing method Methods 0.000 claims description 10
- 238000000034 method Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 5
- 230000002194 synthesizing effect Effects 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 238000003786 synthesis reaction Methods 0.000 claims description 2
- 238000006243 chemical reaction Methods 0.000 claims 1
- 238000003062 neural network model Methods 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000004438 eyesight Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 241001456553 Chanodichthys dabryi Species 0.000 description 1
- 208000029549 Muscle injury Diseases 0.000 description 1
- 206010041591 Spinal osteoarthritis Diseases 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 208000036319 cervical spondylosis Diseases 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000002592 echocardiography Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 208000001491 myopia Diseases 0.000 description 1
- 230000004379 myopia Effects 0.000 description 1
- 208000024335 physical disease Diseases 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 208000005801 spondylosis Diseases 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/52—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00
- G01S7/539—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention belongs to the field of millimeter wave radars, and provides a millimeter wave sitting posture detection intelligent desk lamp based on a convolutional neural network. And (3) utilizing the preset point cloud data under various postures to manufacture a convolutional neural network model, inputting the collected user point cloud data into the trained convolutional neural network model during detection, and matching the user point cloud data with a sitting posture characteristic label in the model to realize the detection of the sitting posture characteristic. The invention can detect the sitting posture characteristics of the user in real time for a long time, has the advantages of strong anti-interference capability and low cost, ensures the degree of freedom of the user during detection, protects the privacy of the user and can bring good use experience to the user.
Description
Technical Field
The application relates to the field of millimeter wave radars, in particular to a millimeter wave sitting posture detection intelligent desk lamp based on a convolutional neural network.
Background
Because the lessons of the teenagers are too heavy, the learning time is longer and longer, and physical diseases and vision problems caused by incorrect sitting postures during learning are more serious and more common, such as humpback, cervical spondylosis, lumbar muscle injury, myopia and the like, which all bring injuries to the teenagers in different degrees. Therefore, to protect the eyesight of teenagers and promote the health of teenagers, the real-time detection of the sitting posture of teenagers during study is an important research subject.
With the increase of the study time of teenagers at night, the desk lamp also becomes an indispensable study article, more and more additional functions are presented on the desk lamp, and besides basic lighting, the multifunctional intelligent desk lamp is endless. The reasonable desk lamp that utilizes combines together the position of sitting detection and wisdom desk lamp, will be the uncomfortable big helping hand of reduction health. The position of sitting of current wisdom desk lamp detects mainly through infrared and camera realization, and the testing result accuracy is higher, but the cost is also higher, easily receives the environmental impact to adopt the camera to detect and probably infringe user's privacy.
Based on the problems, the invention provides the millimeter wave sitting posture detection intelligent desk lamp based on the convolutional neural network, which can detect the sitting posture state of a user in real time, has strong anti-interference capability and also protects the privacy of the user.
Disclosure of Invention
The invention aims to provide a millimeter wave sitting posture detection intelligent desk lamp based on a convolutional neural network.
The technical solution for realizing the invention is as follows: a millimeter wave sitting posture detection method based on a convolutional neural network comprises the following steps:
step one, a millimeter wave radar is placed in a desk lamp, a human body target echo signal is obtained through the radar, the human body target echo signal is preprocessed, the processed human body target echo signal is input into a feature extraction network model trained in advance, and sitting posture features are extracted.
Step two, the human body target echo signal comprises: human target distance, speed, and angle information.
Step three, the preprocessing of the human body target echo signal specifically comprises the following steps:
1) performing Fast Fourier Transform (FFT) on the intermediate frequency signal to obtain a Range curve;
2) determining the range of a target, executing Doppler dimension FFT on a target signal, and acquiring a range-velocity two-dimensional FFT spectrogram;
3) executing angle dimension FFT on a target signal to obtain a distance-angle two-dimensional FFT spectrogram;
4) synthesizing the distance-speed two-dimensional FFT spectrogram and the distance-angle two-dimensional FFT spectrogram into point cloud data;
5) and transforming the point cloud data coordinates into three-dimensional coordinates for display.
And step four, obtaining the distance-speed two-dimensional FFT spectrogram, filtering the indoor static target by using a method of subtracting adjacent frames before and after, and obtaining the distance-speed two-dimensional FFT spectrogram.
And fifthly, taking the three-dimensional coordinates under various postures as the input of the convolutional neural network to train the convolutional neural network.
Step six, the convolutional neural network training comprises the following steps:
1) collecting a plurality of groups of training samples, wherein each group of samples comprises a plurality of user posture characteristics corresponding to the same sitting posture;
2) the training sample group is respectively marked with corresponding labels for classifying different sitting postures;
3) and inputting the training sample into a model for training, namely generating a convolutional neural network model.
And seventhly, converting the three-dimensional coordinates into a picture format, inputting the picture format into the trained convolutional neural network, and judging a group label matched with the posture characteristic of the user by the sitting posture detection model, namely realizing the classification of the sitting posture.
Has the advantages that: compared with the prior art, the technical scheme of the invention has the following beneficial technical effects:
the invention realizes a millimeter wave sitting posture detection intelligent desk lamp based on a convolutional neural network, and is an innovation of the existing sitting posture detection desk lamp. The millimeter wave sitting posture identification result based on the convolutional neural network is more accurate.
In addition, the millimeter wave radar is applied to sitting posture detection, and the advantages of radar anti-interference, small size and high precision are fully exerted.
In addition, the millimeter wave radar is combined with the desk lamp, so that the cost is low, the use is convenient, and the operation is easy.
In addition, compared with an infrared or camera detection method, the non-contact detection method greatly improves the degree of freedom of measurement and protects the privacy of users.
In addition, an idea is provided for the sitting posture detection of the desk lamp, the follow-up precision and the extended functions can be continuously improved, the functions of the desk lamp are more perfect, and the accuracy of the detection result is higher.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.
FIG. 1 is a schematic diagram of a smart desk lamp according to the present invention;
FIG. 2 is a schematic diagram of a millimeter wave sitting posture detection process based on a convolutional neural network provided in the present invention;
fig. 3 is a schematic diagram of a processing flow of a target echo signal provided by the present invention.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.
The invention is further illustrated by the following examples and figures of the specification.
Example 1
A sitting posture detection intelligent desk lamp based on a millimeter wave radar combined convolutional neural network is characterized in that a sitting posture detection process is shown in a figure 1, and the sitting posture detection intelligent desk lamp comprises the following steps:
1) collecting data by a radar: placing a radar in a desk lamp, and carrying out non-contact information acquisition on a human body target by using the radar to transmit frequency-modulated continuous waves;
the radar is a frequency modulation continuous wave radar which transmits frequency modulation continuous waves by utilizing a linear frequency modulation technology;
2) echo signal processing: preprocessing the echo signal of the human target, acquiring distance, speed and angle information of the human target, generating a distance-speed spectrogram and a distance-angle spectrogram, and synthesizing point cloud data;
3) extracting target features: comparing the differences of the point cloud data under different sitting postures, completing calibration, and calibrating the obvious difference characteristics under different sitting postures;
4) making a data set and a test set: and transforming the point cloud data into a three-dimensional coordinate for display, making a data set from the point cloud data in the three-dimensional coordinate in a picture form, and collecting a plurality of groups of data sets as a training set and a test set, wherein each group of training set and test set comprises a plurality of user posture characteristics corresponding to the same sitting posture. The training sets are respectively marked with corresponding labels for classifying different sitting postures;
5) sitting posture recognition, neural network performance analysis and adjustment: inputting the training set into a model for training, matching the obtained sitting posture characteristics with the characteristic group labels through the test set, outputting the corresponding group labels, testing the accuracy of the network, adjusting according to the obtained result, and finishing the classification of the sitting posture.
Example 2
A sitting posture detection intelligent desk lamp based on a millimeter wave radar combined convolutional neural network is disclosed, wherein a target echo signal processing flow is shown in figure 2, and the sitting posture detection intelligent desk lamp comprises the following steps:
1) echo data analysis: analyzing radar echoes frame by frame, completing data extraction on chirp on each antenna dimension of radar original data, and displaying the data in a complex form;
2) distance dimension FFT: after FFT operation is carried out on the intermediate frequency signal, a maximum value is found on a frequency spectrogram, the maximum value is the position of a target, but the horizontal axis is a sampling point at the moment, and the horizontal axis is converted into a distance to obtain a target distance;
3) and (3) FFT of a velocity dimension: searching the position of a peak of a distance spectrogram of each frame of data of the millimeter wave radar, calculating a phase value corresponding to the peak, and obtaining a target speed by using a phase difference between two adjacent frames;
4) angle dimension FFT: performing FFT operation of a time axis on the intermediate frequency signal received by each receiving antenna, finding the frequency of the position of the target of each receiving antenna, performing FFT operation of an antenna axis on the information at the same frequency, finding the maximum value of the second-dimension FFT, obtaining the phase difference of adjacent receiving antennas, and solving the direction of the target;
5) point cloud data synthesis: and synthesizing the distance-speed two-dimensional FFT spectrogram and the distance-angle two-dimensional FFT spectrogram into point cloud data.
Claims (6)
1. The utility model provides a millimeter wave position of sitting detects wisdom desk lamp based on convolutional neural network which characterized in that includes the following step:
step 1: the method comprises the steps of obtaining human body target echo information through a millimeter wave radar, and preprocessing the human body target echo information to obtain target distance, target speed and target angle information.
Step 2: and synthesizing the target distance, the target speed and the target angle information into point cloud data, inputting the point cloud data into a pre-trained feature extraction network, and extracting sitting posture features.
And step 3: and inputting the sitting posture characteristics into a convolution neural network model, and matching the sitting posture characteristics with the existing sitting posture characteristics in the model to realize the classification of the sitting posture characteristics.
2. The method for acquiring the echo information of the human body target through the millimeter wave radar according to the claim 1, and preprocessing the echo information of the human body target to obtain the target distance, the target speed and the target angle information, wherein the method comprises the following steps:
step 1: and acquiring signals by using a frequency modulation continuous wave radar, performing low-pass filtering on the original human body target echo signals, and filtering noise signals in the original human body target echo signals.
Step 2: and sampling the filtered signal, and performing analog-to-digital conversion to obtain a digital signal of the human body target echo.
And step 3: and performing fast Fourier transform (distance dimension FFT) on the human body target echo signal to obtain a Range curve (target distance information).
And 4, step 4: and determining the distance range of the target, and executing Doppler dimension FFT (fast Fourier transform) on the echo signal of the human body target to acquire target speed information.
And 5: and executing angle dimension FFT on the human body target echo signal to acquire target angle information.
3. The object distance information, speed information, and angle information synthesis point cloud data according to claim 1, characterized by comprising the steps of:
step 1: and combining the target distance information and the target speed information to obtain a distance-speed two-dimensional FFT spectrogram.
Step 2: and combining the target distance information and the target angle information to obtain a distance-angle two-dimensional FFT spectrogram.
And step 3: and synthesizing the distance-speed two-dimensional FFT spectrogram and the distance-angle two-dimensional FFT spectrogram into point cloud data.
4. The intelligent desk lamp capable of detecting sitting postures based on the millimeter wave radar combined with the convolutional neural network as claimed in claim 1, wherein point cloud data of various postures are stored into a picture format as an input of a pre-trained convolutional neural network.
5. The pre-trained feature extraction network of claim 1, comprising the steps of:
step 1: and transforming the point cloud data into three-dimensional coordinates for display.
Step 2: and making a data set from the cloud data of the points in the three-dimensional coordinates in a picture form.
And step 3: and collecting a plurality of groups of data sets as a training set and a testing set, wherein each group of training set and testing set comprises a plurality of user posture characteristics corresponding to the same sitting posture.
And 4, step 4: and respectively marking the training sets as corresponding labels for classifying different sitting postures.
And 5: and inputting the training set into a model for training, testing the accuracy of the network through the test set, adjusting according to the obtained result, and generating a feature extraction network.
6. The sitting posture feature input to convolutional neural network model of claim 1, comprising the steps of:
step 1: and inputting the user point cloud data into a feature extraction network to obtain the sitting posture features.
Step 2: and matching the obtained sitting posture characteristics with the characteristic group labels, and outputting the corresponding group labels to finish the sitting posture detection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111448426.XA CN114114223A (en) | 2021-12-01 | 2021-12-01 | Millimeter wave sitting posture detection intelligent desk lamp based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111448426.XA CN114114223A (en) | 2021-12-01 | 2021-12-01 | Millimeter wave sitting posture detection intelligent desk lamp based on convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114114223A true CN114114223A (en) | 2022-03-01 |
Family
ID=80369069
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111448426.XA Pending CN114114223A (en) | 2021-12-01 | 2021-12-01 | Millimeter wave sitting posture detection intelligent desk lamp based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114114223A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115345908A (en) * | 2022-10-18 | 2022-11-15 | 四川启睿克科技有限公司 | Human body posture recognition method based on millimeter wave radar |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110348288A (en) * | 2019-05-27 | 2019-10-18 | 哈尔滨工业大学（威海） | A kind of gesture identification method based on 77GHz MMW RADAR SIGNAL USING |
| CN112034446A (en) * | 2020-08-27 | 2020-12-04 | 南京邮电大学 | Gesture recognition system based on millimeter wave radar |
| CN113050083A (en) * | 2021-03-10 | 2021-06-29 | 中国人民解放军国防科技大学 | Ultra-wideband radar human body posture reconstruction method based on point cloud |
-
2021
- 2021-12-01 CN CN202111448426.XA patent/CN114114223A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110348288A (en) * | 2019-05-27 | 2019-10-18 | 哈尔滨工业大学（威海） | A kind of gesture identification method based on 77GHz MMW RADAR SIGNAL USING |
| CN112034446A (en) * | 2020-08-27 | 2020-12-04 | 南京邮电大学 | Gesture recognition system based on millimeter wave radar |
| CN113050083A (en) * | 2021-03-10 | 2021-06-29 | 中国人民解放军国防科技大学 | Ultra-wideband radar human body posture reconstruction method based on point cloud |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115345908A (en) * | 2022-10-18 | 2022-11-15 | 四川启睿克科技有限公司 | Human body posture recognition method based on millimeter wave radar |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102058411B (en) | UVB based multi-channel radar life detection instrument | |
| Geisheimer et al. | A continuous-wave (CW) radar for gait analysis | |
| Sakamoto et al. | Hand gesture recognition using a radar echo I–Q plot and a convolutional neural network | |
| CN102131290B (en) | WLAN (Wireless Local Area Network) indoor neighbourhood matching positioning method based on autocorrelation filtering | |
| CN113391282B (en) | Human body posture recognition method based on radar multi-dimensional feature fusion | |
| CN110596705B (en) | Human body target identity recognition method and system based on vital sign SAR imaging | |
| CN102073042A (en) | Multichannel multi-human body target two-dimensional positioning method based on UWB (Ultra Wide Band) radar type life detector | |
| CN109407094A (en) | Vehicle-mounted ULTRA-WIDEBAND RADAR forword-looking imaging system | |
| CN114814775A (en) | Radar fall detection method and equipment based on ResNet network | |
| CN106646466B (en) | A kind of imaging method of the weighting back-projection algorithm based on principal component analysis | |
| Qi et al. | Multi-classification algorithm for human motion recognition based on IR-UWB radar | |
| WO2023029390A1 (en) | Millimeter wave radar-based gesture detection and recognition method | |
| CN114114223A (en) | Millimeter wave sitting posture detection intelligent desk lamp based on convolutional neural network | |
| CN112285698B (en) | Multi-target tracking device and method based on radar sensor | |
| Jiang et al. | Recognition of dynamic hand gesture based on mm-wave FMCW radar micro-Doppler signatures | |
| WO2023213051A1 (en) | Static human body posture estimation method based on csi signal angle-of-arrival estimation | |
| CN114814832A (en) | Millimeter wave radar-based real-time monitoring system and method for human body falling behavior | |
| CN116184394A (en) | Millimeter wave radar gesture recognition method and system based on multi-domain spectrogram and multi-resolution fusion | |
| Ding et al. | A fall detection method based on K-nearest neighbor algorithm with MIMO millimeter-wave radar | |
| Qu et al. | Dynamic hand gesture classification based on multichannel radar using multistream fusion 1-D convolutional neural network | |
| CN103027713A (en) | Muscle thickness measuring method and system based on ultrasonic image | |
| CN106726210A (en) | A kind of intelligence control system of multifunctional wheelchair | |
| CN117058228A (en) | Three-dimensional human body posture estimation method based on low-channel radar | |
| Song et al. | High-accuracy gesture recognition using mm-wave radar based on convolutional block attention module | |
| CN115937977A (en) | Few-sample human body action recognition method based on multi-dimensional feature fusion |
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 |