CN110275614B - Non-contact gesture recognition device and method thereof - Google Patents
Non-contact gesture recognition device and method thereof Download PDFInfo
- Publication number
- CN110275614B CN110275614B CN201910460084.XA CN201910460084A CN110275614B CN 110275614 B CN110275614 B CN 110275614B CN 201910460084 A CN201910460084 A CN 201910460084A CN 110275614 B CN110275614 B CN 110275614B
- Authority
- CN
- China
- Prior art keywords
- gesture
- chip microcomputer
- single chip
- section
- data
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/017—Gesture based interaction, e.g. based on a set of recognized hand gestures
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Human Computer Interaction (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
The invention discloses a non-contact gesture recognition device and a method thereof, and the device comprises an FDC2214 capacitance sensing chip, a single-sided copper-clad plate and a single chip microcomputer, wherein the single-sided copper-clad plate and a gesture to be detected form a switch type capacitance, the FDC2214 capacitance sensing chip is electrically connected with the copper plate of the single-sided copper-clad plate and detects and obtains a switch type capacitance value formed by the single-sided copper-clad plate and the gesture to be detected, the output end of the FDC2214 capacitance sensing chip is connected with the single chip microcomputer, and the single chip microcomputer reads frequency data reflecting the switch type capacitance value measured by the FDC2214 capacitance sensing chip; the single chip microcomputer is provided with a memory for storing frequency data, completes classification decision trees of different gestures through training construction by utilizing the frequency data, classifies the gestures of the frequency data acquired in real time by utilizing the classification decision trees and outputs classification results. The method is suitable for recognizing the gestures with different palm sizes, and has high recognition accuracy.
Description
Technical Field
The invention relates to the technical field of gesture recognition, in particular to a non-contact gesture recognition device and a non-contact gesture recognition method.
Background
As shown in fig. 1, FDC2214 is a noise and electromagnetic interference resistant, high resolution, high speed, multi-channel detection sensor based on LC resonant circuit principles, oriented to capacitive sensing solutions, developed by texas instruments, that can achieve non-contact sensing within the sensing region. The FDC2214 detection principle is shown in fig. 1.
The FDC2214 provides 4 detection channels CH 0-CH 3, and the input end of each detection channel is connected with an inductor and a capacitor to form an LC circuit. The sensing end of the measured capacitance (gray mark identification part in fig. 1) is connected with the LC circuit, and an oscillation frequency is generated, and the value of the measured capacitance can be calculated according to the oscillation frequency value.
The series of single-chip microcomputers are 32-bit microcontrollers based on AMR Cortex-M4 cores and are released by Enzhipu company (original flying Secall company), and the series of single-chip microcomputers have the advantages of low power consumption, abundant peripheral modules and the like. The invention adopts MDN512ZVLQ10 singlechip (hereinafter called singlechip) in the series, the singlechip of the type has 512KB flash memory and 128KB random access memory SRAM, which can satisfy the storage of a large amount of gesture data in the training stage and the storage space required when the decision tree construction algorithm is operated. Meanwhile, the single chip microcomputer is provided with various communication interfaces, such as an I2C communication interface, an ethernet MAC controller, a USB communication interface, an SD host controller, a UART communication interface and the like, wherein the I2C communication interface can be used for communicating with the FDC2214, and the UART communication interface, the USB communication interface and the ethernet MAC controller can be used for communicating with a PC, so as to realize the summarization and analysis of gesture data or other application extension functions as required.
Common machine learning methods include K-nearest neighbor, Support Vector Machine (SVM), neural network, deep learning and other algorithms, wherein the K-nearest neighbor algorithm is relatively simple, but when a new sample is identified, a previous learning sample data set needs to be retained, and when a distance is calculated, the distance between the new sample and all samples in the learning sample data set needs to be calculated, so that a large storage overhead and a large calculation overhead are required.
The SVM algorithm is originally designed for a binary classification problem, and for a multi-classification gesture recognition problem, a proper multi-class classifier needs to be constructed, so that the calculation complexity is high, the realization is difficult, and the training speed is relatively slow, so that the SVM algorithm is not an optimal scheme.
For neural networks and deep learning algorithms, a large amount of training data support is required, and a complex parameter adjustment mechanism is required, which is also not an optimal solution for embedded systems with limited storage and computing resources.
Decision tree algorithms are a typical multi-classification approach. The algorithm first processes the data, generates readable rules and decision trees using an inductive algorithm, and then analyzes the new data using the decisions. Decision trees are essentially the process of classifying data through a series of rules. The decision tree algorithm has the advantages of high classification precision, simple generation mode and better robust performance on noise data, and is one of the most widely applied multi-classification algorithms at present. In the training stage, the required training data is not needed to be much, and in the recognition stage, the classification can be obtained by performing comparison for a plurality of times, so that the decision tree algorithm is a better machine learning scheme in terms of storage cost and calculation cost.
Disclosure of Invention
The invention aims to provide a non-contact gesture recognition device and a method thereof, and provides the non-contact gesture recognition device with high cost performance and wide application range so as to accurately and quickly realize a gesture recognition function in an induction area.
The technical scheme adopted by the invention is as follows:
a non-contact gesture recognition device comprises an FDC2214 capacitance sensing chip, a splicing type single-side copper-clad plate and a single chip microcomputer, wherein the splicing type single-side copper-clad plate and a gesture to be detected form a switch type capacitance, the FDC2214 capacitance sensing chip is electrically connected with the copper plate of the splicing type single-side copper-clad plate and is used for detecting and obtaining a switch type capacitance value formed by the splicing type single-side copper-clad plate and the gesture to be detected, the output end of the FDC2214 capacitance sensing chip is connected with the single chip microcomputer, and the single chip microcomputer reads frequency data reflecting the switch type capacitance value measured by the FDC2214 capacitance sensing chip; the single chip microcomputer is provided with a memory for storing frequency data, completes classification decision trees of different gestures by training and constructing through the frequency data, performs gesture classification on the frequency data acquired in real time through the classification decision trees, and outputs a classification result.
Further, the singlechip is a K60 singlechip, and the K60 singlechip is in communication connection with the upper computer through a communication interface arranged on the singlechip.
Further, the communication interface includes a UART communication interface, a USB communication interface, and an ethernet communication interface.
Furthermore, the single-side copper-clad plate is provided with four groups of metal electrodes which can be spliced and are respectively corresponding to the positions of the index finger, the middle finger, the ring finger and the little finger, and whether the index finger, the middle finger, the ring finger and the little finger are present above the metal electrodes is detected, and the four groups of metal electrodes which can be spliced are respectively and electrically connected with 4 capacitance signal acquisition channels on the FDC2214 capacitance sensing chip through a lead.
Furthermore, each group of metal electrodes is of a splicing type and comprises a long-section batten and at least one short-section batten, the long-section batten and the short-section batten are sequentially arranged in a small gap mode along the direction of a corresponding finger, the long-section batten is arranged close to a palm, the short-section batten is arranged far away from the palm, the long-section batten and the short-section batten are connected through a detachable conductive bridge, and when a large palm is detected, more than one short-section batten is spliced on the basis of the long-section batten; when the palm is small, the conductive bridge is disconnected and only long-section battens are used.
Furthermore, the conductive bridge adopts a conductive metal wire or a conductive adhesive.
Furthermore, the small gap interval between the long-section lath and the short-section lath is 1mm, the widths of the long-section lath and the short-section lath are both 2cm, the length of the long-section lath is 10 cm-20 cm, and the length of the short-section lath is 5 cm.
Furthermore, the metal electrodes corresponding to the middle finger and the ring finger are placed in parallel, and the distance between the metal electrodes is 1 cm; the included angles between the metal electrode corresponding to the detection index finger and the metal electrode corresponding to the detection middle finger are all 10 degrees, and the interval of the narrowest part is 0.5 cm; the included angles between the metal electrode corresponding to the ring finger and the metal electrode corresponding to the little finger are both 10 degrees, and the interval of the narrowest position is 0.5 cm.
Further, a non-contact gesture recognition method comprises the following steps:
step 2, the single chip microcomputer respectively collects a plurality of groups of gesture data aiming at different gestures to form a learning data set;
step 3-1, calculating the information entropy of the current learning data set D, wherein the calculation formula is as follows:
wherein p is k Representing the proportion of the kth gesture sample in the n types of gestures in all the learning data sets D;
step 3-2, the frequency data of each group of metal electrodes corresponds to an attribute, and the information gain of all the attributes is calculated by the following calculation formula:
wherein, CHx (x ═ 0,1,2,3), CHx represents channel No. x; d v Is a subset of the learning data set D divided according to different values of CHx, and D ═ D 1 ，D 2 ，…,D V }；
And 3-3, finding out the attribute with the maximum information gain as a branch basis of the current decision tree node, wherein each value of the attribute is used as a child node of the current node.
3-4, repeating the steps 3-1 to 3-3 on all the child nodes until all the child node data sets are gestures of the same type, and further obtaining classification decision trees of different gestures;
and 4, the single chip microcomputer carries out gesture classification on the frequency data acquired in real time by using a classification decision tree and outputs a classification result.
Further, the specific steps of step 2 are:
step 2-1, respectively reading a plurality of groups of frequency data of four channels of FDC2214 to form initial data when the single chip microcomputer is in a gesture-free state;
and 2-2, respectively collecting a plurality of groups of gesture data of each gesture in different application occasions by the single chip microcomputer and forming a learning data set together with the initial data.
Further, the data format of the learning data set in the step 2-2 is stored in the SRAM of the single chip microcomputer according to the format of 'gesture data group' + 'gesture label'.
By adopting the technical scheme, the invention has the following advantages: (1) the non-contact gesture recognition device adopts the FDC2214 capacitive sensor to realize non-contact recognition of gestures, and compared with a computer vision technology recognition scheme, the non-contact gesture recognition device is low in cost and high in environmental adaptability. (2) The spliced metal electrode is suitable for gesture recognition of different palms and has strong individual adaptability. (3) The gesture is recognized by adopting a decision tree algorithm, the accuracy is high, and the time overhead and the storage overhead are small.
Drawings
The invention is described in further detail below with reference to the accompanying drawings and the detailed description;
FIG. 1 is a schematic diagram of the detection of FDC 2214;
fig. 2 is a schematic structural diagram of a non-contact gesture recognition apparatus according to the present invention.
Detailed Description
As shown in fig. 2, the invention discloses a non-contact gesture recognition device, which comprises an FDC2214 capacitance sensing chip, a spliceable single-sided copper-clad plate and a single chip microcomputer, wherein the spliceable single-sided copper-clad plate and a gesture to be detected form a switch type capacitance, the FDC2214 capacitance sensing chip is electrically connected with the spliceable single-sided copper-clad plate, and detects and obtains a switch type capacitance value formed by the spliceable single-sided copper-clad plate and the gesture to be detected, the output end of the FDC2214 capacitance sensing chip is connected to the single chip microcomputer, and the single chip microcomputer reads frequency data reflecting the switch type capacitance value measured by the FDC2214 capacitance sensing chip; the single chip microcomputer is provided with a memory for storing frequency data, completes classification decision trees of different gestures through training construction by utilizing the frequency data, classifies the gestures of the frequency data acquired in real time by utilizing the classification decision trees and outputs classification results.
Specifically, the memory of the K60 single chip microcomputer comprises an SRAM and a FLASH, and the K60 single chip microcomputer stores frequency data into the SRAM and also into the FLASH as required, so that the frequency data can be stored after power failure.
Furthermore, the single-side copper-clad plate is provided with four groups of spliced metal electrodes which respectively correspond to the positions of the index finger, the middle finger, the ring finger and the little finger and detect whether the index finger, the middle finger, the ring finger and the little finger are present above the metal electrodes, and the four groups of metal electrodes are respectively and electrically connected with 4 capacitance signal acquisition channels on the FDC2214 capacitance sensing chip through a lead. In addition, as a possible implementation manner, the gesture collection plane in the hardware solution may adopt a whole single-sided copper-clad plate, and a wire led out from the copper-clad plate is connected to a channel of the FDC2214, so that one FDC2214 can process 4 paths of gesture recognition at the same time.
Furthermore, each group of metal electrodes comprises a long-section batten and at least one short-section batten, the long-section batten and the short-section batten are sequentially arranged with small gaps along the corresponding finger direction, the long-section batten is arranged close to the palm, the short-section batten is arranged far away from the palm, the long-section batten and the short-section batten are connected through a detachable conductive bridge,
further, the singlechip is a K60 singlechip, and the K60 singlechip is in communication connection with the upper computer through a communication interface arranged on the singlechip. Further, the communication interface includes a UART communication interface, a USB communication interface, and an ethernet communication interface. The gesture recognition device can send real-time or historical gesture data to the PC through the communication interfaces for statistical analysis so as to increase the time effectiveness and the space effectiveness of the gesture data.
Specifically, for the gesture sensing plane, because the FDC2214 capacitive sensing chip provides 4 capacitive signal acquisition channels, in order to fully utilize the hardware resources of the FDC2214 and improve the gesture action recognition rate, four groups of metal plates are designed by adopting a single-sided copper-clad plate to be used as fixed electrodes and are sequentially used for detecting whether the index finger, the middle finger, the ring finger and the little finger appear above the metal electrodes.
Because different palms of people are different in size, for example, the palms of adults are usually larger than the palms of children, the palms of men are usually larger than the palms of women, in order to improve the application range of the gesture recognition device, each group of metal electrodes adopts a splicing type, namely two sections of copper-clad battens (hereinafter called battens) with different lengths and the same width are adopted, wherein the length of a long batten can be set to be 10 cm-20 cm according to the size of a common palm, the length of a short batten can be set to be about 5cm, and the width is set to be about 2 cm.
Firstly, two sections of battens are sequentially placed along a certain finger direction, a long section of batten is close to the palm position, a short section of batten is far away from the palm position, the distance between the two sections of battens is as small as possible, but the two sections of battens are not in direct contact, and the two sections of battens are connected as required through a conductive bridge. The conductive bridge can adopt a conductive metal wire or a conductive adhesive. When a large palm is detected, connecting the two sections of battens by using a conductive bridge to be used as a fixed metal electrode plate together; and when a smaller palm is detected, the conductive bridge is disconnected. Long lengths of strip are used in any case, so that one wire is led out to connect with four channels of the FDC2214 acquisition module. Because the fingers are not parallel after being stretched, particularly the little finger and the forefinger have larger stretching angles, the fixed metal electrode groups for measuring the middle finger and the ring finger are placed in parallel, and the distance is set according to the requirement, for example, the distance can be set to be about half of the width of the lath; the electrode group for measuring the index finger and the little finger and the adjacent metal electrode group are arranged according to an included angle of 0-20 degrees, but are not contacted, the narrowest part is one end close to the palm, the distance is set according to the requirement, and for example, the distance can be set to be about half of the width of the fixed metal electrode plate.
Each group of metal electrodes of the gesture sensing plane adopts two sections of single-sided copper-clad plates with the lengths of 10cm and 5cm and the width of 2cm, the distance between long and short strips is 1mm, metal electrode groups of a middle finger and a ring finger are placed in parallel at a distance of 1cm, included angles between metal electrode groups of the index finger and the middle finger and between metal electrode groups of the ring finger and the little finger are 10 degrees, and the distance between the narrowest parts is 0.5 cm.
Further, a non-contact gesture recognition method comprises the following steps:
step 2, the single chip microcomputer respectively collects a plurality of groups of gesture data aiming at different gestures to form a learning data set;
further, the specific steps of step 2 are:
step 2-1, respectively reading a plurality of groups of frequency data of four channels of FDC2214 to form initial data when the single chip microcomputer is in a gesture-free state;
and 2-2, respectively collecting a plurality of groups of gesture data of each gesture by the single chip microcomputer under different application occasions and forming a learning data set together with the initial data. Further, the data format of the learning data set in the step 2-2 is stored in the SRAM of the single chip microcomputer according to the format of 'gesture data group' + 'gesture label'.
Specifically, in the scheme of the gesture recognition device, the data read from the FDC2214 through the single chip microcomputer are frequency data, and the device only needs to distinguish different gestures and does not need to know capacitance values of gesture sensing planes under different gestures, so that the original frequency data read by the device only needs to be directly analyzed.
The frequency data read from the four channels is also different due to the different gestures covering different areas over the four sets of copper clad strips. These four different sets of frequency data can be considered as characteristics of the corresponding gesture.
In the gesture data learning stage, the K60 single chip microcomputer reads a plurality of groups of frequency data of four channels of FDC2214 in an initial state, namely a gesture-free state, the data size is subject to the accurate distinction of different gestures, and for a decision tree algorithm, 100 groups of data are acquired for learning by each gesture, so that the recognition rate of more than 90% can be obtained. And then, sequentially collecting gesture data designed in different application occasions, such as guessing a fist, sequentially collecting three gesture data of 100 groups of 'stone', 'scissors' and 'cloth', and for the finger-punching, sequentially collecting five gesture data of 100 groups of '1', '2', '3', '4' and '5'. The initial data and the gesture data set are combined to form a learning data set. The data format in the learning data set is stored in the SRAM of the K60 single chip microcomputer according to the format of 'gesture data group' + 'gesture label', so that the learning data set is handed to a subsequent decision tree algorithm for learning. And can also be stored in a FLASH memory of the K60 singlechip as required for long-term storage. The saved gesture data can be transmitted to the PC terminal through different communication interfaces according to the requirement.
step 3-1, calculating the information entropy of the current learning data set D, wherein the calculation formula is as follows:
wherein p is k Representing the proportion of the kth gesture sample in the n gestures in all the learning data sets D;
step 3-2, the frequency data of each group of metal electrodes corresponds to an attribute, and the information gain of all the attributes is calculated by the following calculation formula:
wherein, CHx (x ═ 0,1,2,3), CHx represents channel No. x; d v Is a subset of the learning data set D divided according to different values of CHx, and D ═ D 1 ，D 2 ，…,D V }；
And 3-3, finding out the attribute with the maximum information gain as a branch basis of the current decision tree node, wherein each value of the attribute is used as a child node of the current node.
3-4, repeating the steps 3-1 to 3-3 on all the child nodes until all the child node data sets are gestures of the same type, and further obtaining classification decision trees of different gestures;
since the number of commonly used gestures is not large, for example, finger guessing gestures are generally three: the number of the channel data of the stone, the scissors and the cloth is only four, so that the constructed decision tree is usually complicated, the hierarchy is not very deep, the calculation time required by gesture recognition learning based on the classification decision tree is very short, and the method is very suitable for a singlechip application system with relatively limited resources. The decision tree algorithm is realized in a K60 singlechip to obtain a classification decision tree.
And 4, the single chip microcomputer carries out gesture classification on the frequency data acquired in real time by using a classification decision tree and outputs a classification result. After the potential classification decision tree is constructed, the single chip microcomputer collects gesture data in real time and sends the gesture data to the decision tree for processing, classification results can be obtained through comparison for a plurality of times, and the classification results are fed back to a user.
In addition, as a feasible implementation manner, a non-machine learning algorithm can be adopted for the gesture recognition algorithm, and since data of the same gesture are generally close in a certain environment, weighted summation of data of all channels can be adopted to obtain a classification point. The method can be used in an 8-bit singlechip system with weak calculation force and limited memory capacity.
By adopting the technical scheme, the invention is based on a K60 single chip microcomputer and a hardware scheme of a multi-communication-port non-contact gesture recognition device capable of splicing metal electrodes. The gesture recognition algorithm is divided into a learning stage and a recognition stage, a classification decision tree is constructed by using a decision tree algorithm in the learning stage, and gesture classification is carried out by using the constructed classification decision tree in the recognition stage. The invention has the following advantages: (1) the non-contact gesture recognition device adopts the FDC2214 capacitive sensor to realize non-contact recognition of gestures, and compared with a computer vision technology recognition scheme, the non-contact gesture recognition device is low in cost and high in environmental adaptability. (2) The spliced metal electrode is suitable for gesture recognition of different palms and has strong individual adaptability. (3) The gesture is recognized by adopting a decision tree algorithm, the accuracy is high, and the time overhead and the storage overhead are small.
Claims (9)
1. A non-contact gesture recognition method is characterized in that a non-contact gesture recognition device comprises an FDC2214 capacitance sensing chip, a splicing type single-side copper-clad plate and a single chip microcomputer, wherein the splicing type single-side copper-clad plate and a gesture to be detected form a switched capacitor, the FDC2214 capacitance sensing chip is electrically connected with the splicing type single-side copper-clad plate, the switched capacitor formed by the splicing type single-side copper-clad plate and the gesture to be detected is detected and obtained, the output end of the FDC2214 capacitance sensing chip is connected into the single chip microcomputer, and the single chip microcomputer reads frequency data reflecting the switched capacitor measured by the FDC2214 capacitance sensing chip; the single chip microcomputer is provided with a memory for storing frequency data, completes classification decision trees of different gestures through training construction by using the frequency data, performs gesture classification on the frequency data acquired in real time by using the classification decision trees and outputs classification results; the method is characterized in that: the method comprises the following steps:
step 1, four groups of spliced metal electrodes on a single-side copper-clad plate are installed according to the size of a palm to be detected;
step 2, the single chip microcomputer respectively collects a plurality of groups of gesture data aiming at different gestures to form a learning data set;
step 3, training based on the learning data set to obtain classification decision trees of different gestures;
step 3-1, calculating the information entropy of the current learning data set D, wherein the calculation formula is as follows:
wherein p is k Representing the proportion of the kth gesture sample in the n gestures in all the learning data sets D;
step 3-2, the frequency data of each group of metal electrodes corresponds to an attribute, and the information gain of all the attributes is calculated by the following calculation formula:
wherein, CHx (x ═ 0,1,2,3), CHx represents channel No. x; d v Is a subset of the learning data set D divided according to different values of CHx, and D ═ D 1 ，D 2 ，…,D V }；
Step 3-3, finding out the attribute with the maximum information gain as the branch basis of the current decision tree node, and taking each value of the attribute as the child node of the current node;
step 3-4, repeating the steps 3-1 to 3-3 on all the child nodes until all the child node data sets are gestures of the same type, and further obtaining classification decision trees of different gestures;
and 4, the single chip microcomputer carries out gesture classification on the frequency data acquired in real time by using the classification decision tree and outputs a classification result.
2. The non-contact gesture recognition method according to claim 1, wherein: the single chip microcomputer is a K60 single chip microcomputer, and the K60 single chip microcomputer is in communication connection with the upper computer through a communication interface arranged on the K3578 single chip microcomputer.
3. The non-contact gesture recognition method according to claim 2, wherein: the communication interface comprises a UART communication interface, a USB communication interface and an Ethernet communication interface.
4. The non-contact gesture recognition method according to claim 1, wherein: the single-side copper-clad plate is provided with four groups of spliceable metal electrodes, the four groups of spliceable metal electrodes respectively correspond to the positions of the index finger, the middle finger, the ring finger and the little finger and detect whether the index finger, the middle finger, the ring finger and the little finger are present above the metal electrodes, and the four groups of spliceable metal electrodes are respectively and electrically connected with 4 capacitance signal acquisition channels on the FDC2214 capacitance sensing chip through a lead.
5. The method of claim 4, wherein the method comprises: each group of metal electrodes is of a splicing type and comprises a long-section batten and at least one short-section batten, the long-section batten and the short-section batten are sequentially arranged in a small gap mode along the direction of corresponding fingers, the long-section batten is arranged close to a palm, the short-section batten is arranged far away from the palm, the long-section batten and the short-section batten are connected through a detachable conductive bridge, and when a large palm is detected, more than one short-section batten is spliced on the basis of the long-section batten; when detecting a small palm, the conductive bridge is disconnected, and only the long-section batten is used; the conductive bridge adopts a conductive metal wire or a conductive adhesive.
6. The non-contact gesture recognition method according to claim 5, wherein: the small gap interval between the long-section lath and the short-section lath is 1mm, the widths of the long-section lath and the short-section lath are the same and are both 2cm, the length of the long-section lath is 10 cm-20 cm, and the length of the short-section lath is 5 cm.
7. The non-contact gesture recognition method according to claim 5, wherein: the metal electrodes corresponding to the middle finger and the ring finger are arranged in parallel, and the distance between the metal electrodes is 1 cm; the included angle between the metal electrode for correspondingly detecting the index finger and the metal electrode for detecting the middle finger is 10 degrees, and the interval of the narrowest part is 0.5 cm; the included angle between the metal electrode corresponding to the ring finger and the metal electrode corresponding to the little finger is 10 degrees, and the interval of the narrowest part is 0.5 cm.
8. The non-contact gesture recognition method according to claim 1, wherein: the specific steps of the step 2 are as follows:
step 2-1, respectively reading a plurality of groups of frequency data of four channels of FDC2214 to form initial data when the single chip microcomputer is in a gesture-free state;
and 2-2, respectively collecting a plurality of groups of gesture data of each gesture in different application occasions by the single chip microcomputer and forming a learning data set together with the initial data.
9. The non-contact gesture recognition method according to claim 1, wherein: and (3) storing the data format of the learning data set in the step (2-2) into an SRAM of the singlechip according to the format of the gesture data group plus the gesture label.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910460084.XA CN110275614B (en) | 2019-05-30 | 2019-05-30 | Non-contact gesture recognition device and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910460084.XA CN110275614B (en) | 2019-05-30 | 2019-05-30 | Non-contact gesture recognition device and method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110275614A CN110275614A (en) | 2019-09-24 |
| CN110275614B true CN110275614B (en) | 2022-09-30 |
Family
ID=67960378
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910460084.XA Active CN110275614B (en) | 2019-05-30 | 2019-05-30 | Non-contact gesture recognition device and method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110275614B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112099624B (en) * | 2020-08-25 | 2022-11-22 | 李志斌 | Multimode diamond-shaped frame type capacitive sensing gesture recognition system |
| CN113673292A (en) * | 2021-01-14 | 2021-11-19 | 南方科技大学 | Capacitive imaging sensor and gesture form sensing method |
| CN114327054A (en) * | 2021-12-21 | 2022-04-12 | 杭州电子科技大学 | Gesture recognition device based on FDC2214 |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN203287847U (en) * | 2013-05-31 | 2013-11-13 | 深圳职业技术学院 | Gesture recognition all-in-one machine based on MGC3130 chip |
| KR20150030072A (en) * | 2013-09-11 | 2015-03-19 | 삼성전기주식회사 | Touch sensor to recognize a gesture and controlling method thereof |
| KR20170091963A (en) * | 2016-02-02 | 2017-08-10 | 삼성전자주식회사 | Gesture classification apparatus and method using electromyogram signals |
| CN107526440A (en) * | 2017-08-28 | 2017-12-29 | 四川长虹电器股份有限公司 | The intelligent electric appliance control method and system of gesture identification based on decision tree classification |
| CN109032349A (en) * | 2018-07-10 | 2018-12-18 | 哈尔滨工业大学 | A kind of gesture identification method and system based on millimetre-wave radar |
| CN109343694A (en) * | 2018-08-13 | 2019-02-15 | 浙江大学 | A kind of gesture recognition system and method for finger-guessing game finger-guessing game game |
| CN109491507A (en) * | 2018-11-14 | 2019-03-19 | 南京邮电大学 | Gesture identifying device based on FDC2214 |
| CN109710116A (en) * | 2018-08-23 | 2019-05-03 | 华东师范大学 | A kind of non-contact gesture state recognition system and recognition methods |
| CN109799914A (en) * | 2019-03-18 | 2019-05-24 | 大连理工大学 | A kind of gesture identifying device based on FDC2214 |
-
2019
- 2019-05-30 CN CN201910460084.XA patent/CN110275614B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN203287847U (en) * | 2013-05-31 | 2013-11-13 | 深圳职业技术学院 | Gesture recognition all-in-one machine based on MGC3130 chip |
| KR20150030072A (en) * | 2013-09-11 | 2015-03-19 | 삼성전기주식회사 | Touch sensor to recognize a gesture and controlling method thereof |
| KR20170091963A (en) * | 2016-02-02 | 2017-08-10 | 삼성전자주식회사 | Gesture classification apparatus and method using electromyogram signals |
| CN107526440A (en) * | 2017-08-28 | 2017-12-29 | 四川长虹电器股份有限公司 | The intelligent electric appliance control method and system of gesture identification based on decision tree classification |
| CN109032349A (en) * | 2018-07-10 | 2018-12-18 | 哈尔滨工业大学 | A kind of gesture identification method and system based on millimetre-wave radar |
| CN109343694A (en) * | 2018-08-13 | 2019-02-15 | 浙江大学 | A kind of gesture recognition system and method for finger-guessing game finger-guessing game game |
| CN109710116A (en) * | 2018-08-23 | 2019-05-03 | 华东师范大学 | A kind of non-contact gesture state recognition system and recognition methods |
| CN109491507A (en) * | 2018-11-14 | 2019-03-19 | 南京邮电大学 | Gesture identifying device based on FDC2214 |
| CN109799914A (en) * | 2019-03-18 | 2019-05-24 | 大连理工大学 | A kind of gesture identifying device based on FDC2214 |
Non-Patent Citations (1)
| Title |
|---|
| "一种非接触式的手势识别装置";罗于恒，王洋，刘伟;《科技与创新》;20181115;第55-56页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110275614A (en) | 2019-09-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110275614B (en) | Non-contact gesture recognition device and method thereof | |
| CN107633227B (en) | CSI-based fine-grained gesture recognition method and system | |
| Li et al. | Two-stream convolution augmented transformer for human activity recognition | |
| CN105844216B (en) | Detection and matching mechanism for recognizing handwritten letters by WiFi signals | |
| Chen et al. | Hand gesture recognition using Haar-like features and a stochastic context-free grammar | |
| CN103257711A (en) | Space gesture input method | |
| CN102867190B (en) | A kind of method utilizing mobile device built-in sensors to carry out Activity recognition | |
| CN110674875A (en) | Pedestrian motion mode identification method based on deep hybrid model | |
| CN112380976A (en) | Gesture recognition system and method based on neural network visual touch sensor fusion | |
| Lin et al. | WiWrite: An accurate device-free handwriting recognition system with COTS WiFi | |
| CN106406516A (en) | Local real-time movement trajectory characteristic extraction and identification method for smartphone | |
| CN109343694B (en) | Gesture recognition system and method for finger-guessing and punching game | |
| Anupama et al. | Automated sign language interpreter using data gloves | |
| CN104680142A (en) | Method for comparing four-slap fingerprint based on feature point set segmentation and RST invariant features | |
| Zhang et al. | Wi-adaptor: Fine-grained domain adaptation in wifi-based activity recognition | |
| CN116400831B (en) | Handwriting pressure induction compensation algorithm of passive electromagnetic touch screen | |
| CN109710116B (en) | Non-contact gesture state recognition system and recognition method | |
| CN102523091A (en) | Intelligent authentication system and method based on wireless power supply sensor | |
| CN113300750A (en) | Personnel identity authentication and handwritten letter identification method based on WIFI signal | |
| Darafsh et al. | Real-Time activity recognition and intention recognition using a vision-based embedded system | |
| CN109283998A (en) | Three-dimensional capacitive wearable human-computer interaction device and method | |
| CN202334554U (en) | Novel passive intelligent identification system | |
| CN111160391A (en) | Space division-based rapid relative density noise detection method and storage medium | |
| WO2019193601A1 (en) | Piezoelectric sensors integrated hand gesture recognition device | |
| US20230015735A1 (en) | Capacitive sensor, electronic device, and electronic device control 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 | ||
| GR01 | Patent grant |