CN116149475A

CN116149475A - Gesture recognition method and device, equipment and computer readable storage medium

Info

Publication number: CN116149475A
Application number: CN202310073666.9A
Authority: CN
Inventors: 张溶冰
Original assignee: Suzhou Xinrong Radium Technology Partnership LP
Current assignee: Suzhou Xinrong Radium Technology Partnership LP
Priority date: 2023-01-20
Filing date: 2023-01-20
Publication date: 2023-05-23

Abstract

A gesture recognition method and device, equipment and a computer readable storage medium, wherein the gesture recognition method comprises the following steps: detecting a first trigger signal; determining whether a possible gesture action is found according to the first trigger signal; the first trigger signal is triggered by temperature jump of a target induction area sensed by the temperature sensor. By adopting the scheme, the gesture recognition precision can be improved.

Description

Gesture recognition method and device, equipment and computer readable storage medium

Technical Field

The present invention relates to the field of intelligent electronic devices, and in particular, to a gesture recognition method and apparatus, a device, and a computer readable storage medium.

Background

With the development of electronic technology, wearable intelligent devices, such as smart watches, smart bracelets, smart glasses, etc., are widely used in daily life. Compared with the traditional control mode of touching the intelligent equipment by fingers, the non-contact gesture recognition technology can control the intelligent equipment by monitoring gesture actions of a user, and non-contact control of the intelligent equipment can be realized.

In the prior art, non-contact gesture recognition technologies include optical, capacitive, radio frequency, infrared LED, microwave, video image, and the like. However, the existing non-contact recognition technology recognizes module power consumption and volume limitation, or is affected by external ambient light and ambient temperature fluctuation in the recognition process, so that the method is not suitable for wearable devices.

Disclosure of Invention

The embodiment of the invention adopts the passive infrared temperature sensor, and realizes the accurate identification with low power consumption, small volume and under various environments through system hardware and algorithm.

In order to solve the above technical problems, an embodiment of the present invention provides a gesture recognition method, including: detecting a first trigger signal; determining whether a possible gesture action is found according to the first trigger signal; the first trigger signal is triggered by target induction area temperature jump induced by the temperature sensor.

Optionally, the determining whether to find a possible gesture according to the first trigger signal includes: determining that the temperature of the target induction area jumps according to the first trigger signal; and determining that gesture actions exist when the frequency of the jump of the temperature of the target sensing area within the preset time is larger than a preset value.

Optionally, when the number of times that the temperature of the target sensing area jumps within the preset time period is greater than a preset value, determining that a gesture action exists includes: and when continuous jump of the temperature of the target sensing area within the preset time period is detected and a certain trend exists, determining that gesture actions exist.

Optionally, the detecting that the temperature of the target sensing area continuously jumps within the preset time period includes: the trend direction of detecting the continuous temperature jump of the target induction area is opposite.

Optionally, the detecting the first trigger signal includes: filtering the received signal of the temperature sensor to obtain a filtered signal; the first trigger signal is detected from the filtered signal.

Optionally, the filtering the received signal of the temperature sensor includes: carrying out weighted average on the received signal and the latest N times of historical received signals, and taking an average result as the filtering signal; n is a positive integer.

Optionally, the detecting the first trigger signal from the filtered signal includes any one of the following: when the deviation between the filtering signal and the adjacent latest M historical filtering signals is larger than a first deviation value, determining that the first triggering signal is detected, wherein M is a positive integer; or determining that the first trigger signal is detected when the deviation between the filtered signal and the adjacent latest M historical filtered signals is larger than the first deviation value and the deviation between the filtered signal and other historical filtered signals is larger than the second deviation value; the other historical filtered signals are not adjacent to the last M historical filtered signals; or, the deviation between the filtering signal and the X times of historical filtering signals is larger than the first deviation value, and the M times of historical filtering signals are in monotonically increasing relation, so that the first trigger signal is determined to be detected; the X times of historical filtering signals are selected from the M times of historical filtering signals; or performing Fourier frequency domain transformation on the filtered signal and the M times of historical filtered signals, and determining that the first trigger signal is detected if a multi-order frequency offset component exists in a frequency domain transformation result.

Optionally, the gesture recognition method further includes: determining the first deviation value and/or the second deviation value according to the temperature difference between the reference human body temperature and the ambient temperature; the first deviation value is positively correlated with the temperature difference; the second deviation value is positively correlated to the temperature difference.

Optionally, the gesture recognition method further includes: detecting whether an optical window of the temperature sensor is polluted; if the light window is detected to be polluted, generating and outputting reminding information; the reminding information is used for indicating that the light window is polluted.

Optionally, the detecting whether the optical window is contaminated includes: acquiring the current ambient temperature; and if the difference between the current ambient temperature obtained by the temperature sensor and the ambient temperature obtained by the temperature sensor obtained by statistics is larger than a preset difference, determining that the optical window is polluted.

The embodiment of the invention also provides a gesture recognition device, which comprises: the detection unit is used for detecting the first trigger signal; the determining unit is used for determining whether possible gesture actions are found according to the first trigger signal; the first trigger signal is triggered by target induction area temperature jump induced by the temperature sensor.

The embodiment of the invention also provides gesture recognition equipment, which comprises: temperature sensor, amplifier circuit and microprocessor, wherein: the temperature sensor is suitable for detecting the temperature change of the target induction area, generating a temperature change signal and outputting the temperature change signal to the amplifying circuit; the amplifying circuit is suitable for amplifying the temperature change signal and outputting the amplified temperature change signal to the microprocessor; and the microprocessor is suitable for determining whether gesture actions exist according to the jump condition of the temperature change signal within the preset duration.

Optionally, the temperature sensor is a non-contact temperature sensor, including any one of the following: an infrared thermopile sensor, a PIR infrared human body sensor and a vanadium oxide temperature sensor.

The embodiment of the invention also provides a computer readable storage medium, which is a non-volatile storage medium or a non-transient storage medium, and a computer program is stored on the computer readable storage medium, and the computer program is executed by a processor to execute the steps of any gesture recognition method.

Compared with the prior art, the technical scheme of the embodiment of the invention has the following beneficial effects:

and detecting a first trigger signal, and determining whether gesture actions exist according to the first trigger signal. Because the first trigger signal is triggered by the temperature jump of the target sensing area sensed by the temperature sensor, the temperature change of the target sensing area can be judged by detecting the first trigger signal, and whether the possible gesture action is found or not is judged according to the temperature change condition, so that the accuracy of gesture action recognition can be improved.

Further, the received signal of the temperature sensor is subjected to filtering processing, the received signal and the last N times of historical received signals are subjected to weighted average, and the obtained result is used as a filtering signal, so that the obtained filtering signal can eliminate interference of factors such as noise as much as possible, and further the misjudgment probability is reduced. The filtered signal is compared with the latest M times of historical filtered signals to detect whether the first trigger signal exists or not, so that the influence caused by the change of the ambient temperature can be reduced as much as possible, and the gesture recognition accuracy is further improved.

In addition, the hand waving direction corresponding to the gesture motion is obtained, so that flexible control of the intelligent device can be realized.

The current environment temperature is obtained, and whether the optical window is polluted or not is determined according to comparison between the current environment temperature and the environment temperature obtained through statistics, so that misjudgment caused by pollution of the optical window can be effectively avoided.

Drawings

FIG. 1 is a schematic diagram of a gesture recognition apparatus in an embodiment of the present invention;

FIG. 2 is a flow chart of a gesture recognition method in an embodiment of the present invention;

fig. 3 is a schematic structural diagram of a gesture recognition apparatus according to an embodiment of the present invention.

Detailed Description

As described in the background art above, the existing non-contact recognition technology has low recognition accuracy.

In the embodiment of the invention, since the first trigger signal is triggered by the temperature jump of the target sensing area sensed by the temperature sensor, the temperature change of the target sensing area can be judged by detecting the first trigger signal, and whether the possible gesture action is found or not can be judged according to the temperature change condition, so that the accuracy of gesture action recognition can be improved.

In order to make the above objects, features and advantages of the present invention more comprehensible, embodiments accompanied with figures are described in detail below.

Referring to fig. 1, a schematic structural diagram of a gesture recognition apparatus in an embodiment of the present invention is given. In an embodiment of the present invention, the gesture recognition apparatus may include a temperature sensor 101, an amplifying circuit, a microprocessor 103, a power source (not shown in fig. 1), and the like. The gesture recognition device may be built into the smart device.

In the embodiment of the present invention, the temperature sensor 101 may acquire the change of the temperature of the target sensing area, generate a temperature change signal, and output the temperature change signal to the amplifying circuit. The temperature change signal may be characterized in the form of a potential difference signal. The target sensing area may be an operating area of the temperature sensor 101. In other words, the temperature sensor 101 can acquire the temperature within the target sensing region.

In implementations, the temperature sensor 101 may be a non-contact temperature sensor. Specifically, the temperature sensor may be an infrared thermopile sensor, or an infrared human body sensor (PIR), and may also be a vanadium oxide temperature sensor. It should be understood that the specific type of the non-contact temperature sensor is not limited to the above exemplary description, and will not be described herein.

In one embodiment of the invention, the temperature sensor is an infrared thermopile sensor. The infrared thermopile sensor comprises a sensor circuit formed by connecting a plurality of thermocouples in series, and can be installed in intelligent equipment.

In an embodiment of the present invention, the amplifying circuit may include an operational amplifying circuit 102. To improve the amplification accuracy, the operational amplifier circuit 102 may be a high-accuracy operational amplifier circuit. The processor may be an embedded processor chip.

The specific structure and working principle of the gesture recognition apparatus provided in the embodiment of the present invention are described below.

The amplifying circuit may include an operational amplifying circuit 101, a first resistor R1, and a second resistor R2. The first resistor R1 and the second resistor R2 may be high-precision resistors, the resistance of the first resistor R1 may be 1kΩ, and the resistance of the second resistor R2 may be 510kΩ.

When there is a temperature difference between the intelligent device and the ambient temperature, the temperature sensor 101 may generate a corresponding temperature change signal (such as a potential difference signal), and the signal is superimposed with the reference voltage VREF and then used as an input of the positive terminal (+) of the operational amplifier circuit 102. The negative terminal (—) of the operational amplifier circuit 102 inputs a reference voltage VREF, which may be generated and output by a power supply. The power supply may include a reference voltage generating device for generating a reference voltage VREF independent of the power supply voltage.

The amplifying circuit amplifies the potential difference signal generated by the temperature sensor 101 and outputs the amplified potential difference signal to the processor 103. The processor 103 may first analog-to-digital convert the amplified potential difference signal to a digital signal. In the process of carrying out analog-digital conversion on the amplified potential difference signals, the working principle of successive comparison can be adopted, and the sampling rate can reach high-speed sampling of millions of times per second.

The processor 103 may perform corresponding processing on the digital signal, determine whether a gesture exists according to the digital signal, and output a determination result.

In implementations, the gesture recognition device may also include a lens (not shown in fig. 1), which may include a lens. The lens can transmit the infrared light in the target area ranging from 1cm to 30cm (note: concave lens is adopted in most cases), or transmit the infrared light in the target area ranging from 20cm to 6 m, and the infrared light with the angle of 0-55 degrees with the central line of the receiving cavity to the corresponding thermopile induction area.

The specific operation of the processor 103 will be described below. Referring to fig. 2, a gesture recognition method in an embodiment of the present invention is provided. In an embodiment of the present invention, the gesture recognition method described below may be executed by the processor 103.

In step 201, a first trigger signal is detected.

In implementations, the processor may receive a received signal of the temperature sensor. Specifically, the received signal may be the above-mentioned digital signal, that is, the amplified potential difference signal after the analog-to-digital conversion.

In implementations, the temperature sensor may detect the target sensing region in real time or periodically. For example, the temperature sensor detects the target sensing region with a period of 0.2s. Therefore, the temperature sensor detects the temperature of the target sensing region substantially continuously or periodically (due to the short period, it can be regarded as uninterrupted), and thus continuously outputs the reception signal. The processor may detect the first trigger signal therefrom after acquiring the received signal of the temperature sensor.

In a specific implementation, due to the influence of noise and other factors, a certain error may exist in the received signal input to the processor. In order to improve the accuracy of gesture recognition, in the embodiment of the invention, the processor may perform filtering processing on the received signal to obtain a filtered signal; acquiring the jump condition of the filtering signal; and detecting the first trigger signal according to the jump condition of the filtering signal.

The received signal is filtered, the received signal and the last N times of historical received signals can be weighted averaged, and the average result is used as a filtered signal.

In an embodiment of the present invention, the processor may store the received signal. And compared with the currently received signal, the last received signal received by the processor is the historical received signal. The processor may store the last N historical received signals.

For example, when the processor acquires the received signal at both time t0 and time t1, the received signal at time t0 is the historical received signal with respect to time t 1.

In a specific implementation, the value of N may also be set according to specific application requirements. If the accuracy requirement for gesture recognition is high, N may take a large value, such as n=10. If the accuracy requirement for gesture recognition is low, N may take a small value, such as n=5.

In an embodiment of the present invention, after the filtered signal is obtained, the filtered signal may be compared with the last M times of historical filtered signal. And if the deviation between the filtered signal and the latest M times of historical filtered signals is larger than a preset first deviation value, determining that the first trigger signal is detected.

In implementations, the processor may store each acquired filtered signal. The previously stored filtered signal may be considered a historical filtered signal with respect to the current time instant. For example, when the processor obtains the filtered signal at both time t0 and time t1, the filtered signal obtained at time t0 is the historical filtered signal with respect to time t 1.

In the embodiment of the invention, the filtered signal and the historical filtered signal can be presented in the form of digital signals, and the voltage amplitude corresponding to the filtered signal and the historical filtered signal is represented. The deviation between the filtered signal and the historical filtered signal is substantially the deviation between the voltage amplitude corresponding to the filtered signal and the voltage amplitude corresponding to the historical filtered signal.

Thus, the filtered signal is subtracted from the last M historical filtered signals to obtain a difference between the filtered signal and the last M historical filtered signals. When the absolute values of the M differences (i.e., the deviation between the two) are all greater than the preset first deviation value, it may be determined that the first trigger signal is detected.

In a specific implementation, the value of M may also be set according to specific application requirements. If the accuracy requirement for gesture recognition is high, M may take a large value, such as m=4. If the accuracy requirement for gesture recognition is low, M may take a small value, such as m=2.

In the embodiment of the invention, after the filtered signal is obtained, the filtered signal can be compared with the adjacent latest M times of historical filtered signals. If the deviation between the filtered signal and the last M historical filtered signals is larger than the preset first deviation value, and the deviation between the filtered signal and other historical filtered signals is larger than the second deviation value, the first trigger signal can be determined to be detected. The other history filter signals are history filter signals which are different from the M times of history filter signals and are not adjacent to each other.

For example, in chronological order, the 5 times of historical filtered signals adjacent to the filtered signal are in order: a filtered signal 5, a filtered signal 4, a filtered signal 3, a filtered signal 2 and a filtered signal 1. Namely: the filtered signal 1 is a history filtered signal that is most adjacent in time sequence to the filtered signal. Setting m=2. The filtered signal is compared with the filtered signal 1, the filtered signal 2 and the filtered signal 4. If the deviation between the filtering signal and the filtering signal 1 and the deviation between the filtering signal 2 are larger than the first deviation value, and the deviation between the filtering signal and the filtering signal 3 is larger than the second deviation value, the first trigger signal is detected.

In the embodiment of the invention, after the filtered signal is obtained, the X times of historical filtered signals can be selected from the adjacent latest M times of historical filtered signals, and the filtered signal is compared with the X times of historical filtered signals. And if the deviation between the filtered signal and the X historical filtered signals is larger than the first deviation value, and the M historical filtered signals are in monotonically increasing relation, determining that the first trigger signal is detected. X is a positive integer and X is more than or equal to 1 and less than or equal to M.

For example, in chronological order, the 5 times of historical filtered signals adjacent to the filtered signal are in order: a filtered signal 5, a filtered signal 4, a filtered signal 3, a filtered signal 2 and a filtered signal 1. M=5, x=2. The voltage amplitudes corresponding to the filter signal 5, the filter signal 4, the filter signal 3, the filter signal 2 and the filter signal 1 are monotonically increased, that is, the voltage amplitude corresponding to the filter signal 1 is the largest, the voltage amplitude corresponding to the filter signal 2 is the second, and the like, and the voltage amplitude corresponding to the filter signal 5 is the smallest. And if the deviation of the filtered signal, the filtered signal 5 and the filtered signal 4 is larger than the first deviation value, determining that the first trigger signal is detected.

In the embodiment of the present invention, after the filtered signal is obtained, fourier frequency domain transformation may be performed on the filtered signal and the M times of history filtered signal. If the frequency deviation component with the second order and above exists in the frequency domain transformation result (namely, a multi-order component exists), the first trigger signal is detected.

In a specific implementation, the first deviation value and the second deviation value may be different. In implementations, the first deviation value may be less than the second deviation value.

In an embodiment of the present invention, the first deviation value may be dynamically changed. Specifically, the first deviation value may be determined according to a temperature difference between the ambient temperature and the reference human body temperature. The first deviation value may be in a positive correlation with the temperature difference. The reference body temperature may be preset to be a fixed value. For example, the reference human body temperature is set to 36.5 ℃.

The positive correlation between the first deviation and the temperature difference may be: the temperature difference between the ambient temperature and the reference human body temperature is larger, and the first deviation value can take a larger value; the temperature difference between the ambient temperature and the reference human body temperature is small, and the first deviation value may take a small value.

For example, the ambient temperature is 25 ℃, the reference human body temperature is 36.5 ℃, and the first deviation value is 0.5V. The ambient temperature is 30 ℃, the reference human body temperature is 36.5 ℃, and the first deviation value is 0.3V.

The second deviation value may also be determined according to a temperature difference between the ambient temperature and the reference human body temperature, and the second deviation value and the temperature difference may also be in a positive correlation.

In the embodiment of the invention, the first deviation value and the second deviation value are positively correlated with the temperature difference. When the ambient temperature is relatively close to the reference human body temperature, the accuracy can be improved by adjusting the first deviation value and the second deviation value and using smaller first deviation value and second deviation value.

Step 202, determining whether a possible gesture motion is found according to the first trigger signal.

In specific implementation, whether the temperature of the target induction area jumps or not can be determined according to the first trigger signal; if the frequency of the jump of the temperature of the target sensing area is larger than a preset value within a preset time length, determining that gesture actions exist; otherwise, if the frequency of the jump of the temperature of the target sensing area in the preset time period is smaller than the preset value, determining that no gesture motion exists.

In the embodiment of the invention, when the temperature of the target sensing area jumps, correspondingly, the receiving signal of the temperature sensor likewise jumps. In other words, when the temperature of the target sensing area jumps, the first trigger signal exists in the received signal correspondingly.

Conversely, when the first trigger signal is present in the received signal, this means that the target sensing area temperature jumps. Therefore, by detecting the first trigger signal, whether the temperature of the target induction area jumps or not and the frequency of jumping can be obtained. And when the frequency of the jump of the temperature of the target sensing area within the preset time is greater than a preset value, determining that gesture actions exist.

In a specific implementation, the preset duration may be set according to a specific application scenario. The predetermined duration may be a verification value. In one embodiment of the present invention, the preset time period is 0.2s. In another embodiment of the present invention, the preset time period is 1s. It can be understood that the value of the preset duration does not affect the protection scope of the application.

The preset value can also be set according to a specific application scene. In the embodiment of the invention, if continuous jump of the temperature of the target sensing area within the preset time period is detected and a certain trend exists in the continuous jump, gesture motion can be determined. Further, if the trend direction of the continuous temperature jump of the target sensing area is detected to be opposite, determining that a gesture action exists.

In practical application, if the palm of the user swings from above the gesture recognition device, the swing process is as follows: the palm is gradually connected with the near infrared thermopile sensor and then gradually far away from the infrared thermopile sensor. When the palm is gradually connected with the near infrared thermopile sensor, the temperature of the target induction area is gradually increased; as the palm is progressively farther from the infrared thermopile sensor, the target sensing region temperature is progressively lower.

Therefore, within the preset time, the direction of detecting the temperature jump of the target sensing area is the temperature rise, and then the direction of detecting the temperature jump of the target sensing area is the temperature fall, and the gesture action is determined.

In the embodiment of the invention, when the difference between the ambient temperature and the reference human body temperature is smaller, the gesture recognition equipment can be suspended to work so as to avoid misoperation caused by misrecognition.

For example, the ambient temperature is 36 ℃, the reference human body temperature is 36.5 ℃, and the difference between the two is only 0.5 ℃. At this time, the gesture recognition apparatus has a high misrecognition. To avoid misoperations caused by misidentification, the gesture recognition device may be suspended from operation.

In implementations, the light window of the temperature sensor may be contaminated during use of the smart device. For example, when a user uses a smart device, a finger touches the light window, causing oil stains on the light window. If the gesture recognition apparatus is continued to be used, the gesture recognition result (i.e., whether or not there is a gesture motion) obtained by the final recognition may be erroneous.

In order to avoid misjudgment of the gesture recognition result, in the embodiment of the invention, if the optical window is detected to be polluted, reminding information can be generated and output to the intelligent device, and the reminding information can be used for indicating that the optical window is polluted. After the intelligent device receives the reminding information, the intelligent device can remind the user that the optical window is polluted, and the user can clean the optical window.

In the embodiment of the invention, if stains exist on the optical window, the current environmental temperature acquired by the temperature sensor may be abnormal. The temperature sensor can acquire the ambient temperature at regular time, and if the difference between the current ambient temperature and the ambient temperature acquired last time is detected to be larger than a preset value, the light window can be determined to be polluted.

For example, the temperature sensor obtains the current ambient temperature to be 25 ℃, and the last obtained ambient temperature to be 35 ℃, the preset value to be 5 ℃, at this time, it is determined that the light window is contaminated.

In a specific implementation, after the light window is determined to be polluted, the gesture recognition device can be suspended to work so as to avoid misoperation caused by misrecognition.

In implementations, the recognition result may be output to the smart device after determining that a gesture action is present or after determining that no gesture action is present. The intelligent device can correspondingly control the installed application software according to the identification result.

For example, when the recognition result is that the gesture action exists, the intelligent device controls the music playing software to play the next song.

For another example, when the recognition result is that the gesture action exists, the intelligent device controls the browsing interface of the browser to enter the next page.

In a specific implementation, the gesture recognition device may further obtain a hand waving direction corresponding to the gesture motion, and output the hand waving direction. The intelligent device can correspondingly control the installed application software according to the hand swing direction.

For example, when the recognition result is that a gesture motion exists and the waving direction is from left to right, the intelligent device controls the music playing software to play a next song; when the hand waving direction is from right to left, the intelligent device controls the music playing software to pause playing.

In a specific implementation, the triggered sequence of different thermocouples in the infrared thermopile sensor can be obtained, so that the waving direction is determined.

In embodiments of the present invention, the gesture recognition device may further include a plurality of infrared thermopile sensors. Different infrared thermopile sensors are distributed at different locations. And determining the waving direction by acquiring the sequence of gesture actions detected by different infrared thermopile sensors.

For example, the gesture recognition device includes two infrared thermopile sensors, a first infrared thermopile sensor being disposed to the left of a second infrared thermopile sensor. The first infrared thermopile sensor detects gesture motion first, and the second infrared thermopile sensor detects gesture motion later, so that the hand waving direction of the user is determined to be from left to right.

After the existence of the gesture action is determined, the hand waving direction corresponding to the gesture action is identified, and further more flexible control of the intelligent equipment can be achieved.

Referring to fig. 3, a gesture recognition apparatus 30 according to an embodiment of the present invention is provided, including: a detection unit 301 and a determination unit 302, wherein:

a detection unit 301 for detecting a first trigger signal;

a determining unit 302, configured to determine whether a possible gesture is found according to the first trigger signal; the first trigger signal is triggered by target induction area temperature jump induced by the temperature sensor.

In specific implementation, the specific execution process of the detecting unit 301 and the determining unit 302 may refer to the steps 201 to 202 correspondingly, which is not described herein.

Those of ordinary skill in the art will appreciate that all or a portion of the steps in the various methods of the above embodiments may be implemented by a program that instructs related hardware, the program may be stored on a computer readable storage medium, and the storage medium may include: ROM, RAM, magnetic or optical disks, etc.

Although the present invention is disclosed above, the present invention is not limited thereto. Various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention, and the scope of the invention should be assessed accordingly to that of the appended claims.

Claims

1. A method of gesture recognition, comprising:

detecting a first trigger signal;

determining whether a possible gesture action is found according to the first trigger signal; the first trigger signal is triggered by target induction area temperature jump induced by the temperature sensor.

2. The gesture recognition method of claim 1, wherein the determining whether a possible gesture motion is found according to the first trigger signal comprises:

determining that the temperature of the target induction area jumps according to the first trigger signal;

and determining that gesture actions exist when the frequency of the jump of the temperature of the target sensing area within the preset time is larger than a preset value.

3. The gesture recognition method of claim 2, wherein determining that a gesture exists when the number of times the temperature jump of the target sensing area occurs within the preset time period is greater than a preset value comprises:

and when continuous jump of the temperature of the target sensing area within the preset time period is detected and a certain trend exists, determining that gesture actions exist.

4. The gesture recognition method according to claim 3, wherein the detecting that the target sensing area temperature continuously jumps within the preset time period includes:

the trend direction of detecting the continuous temperature jump of the target induction area is opposite.

5. The gesture recognition method of claim 1, wherein detecting the first trigger signal comprises:

filtering the received signal of the temperature sensor to obtain a filtered signal;

the first trigger signal is detected from the filtered signal.

6. The gesture recognition method of claim 5, wherein the filtering the received signal of the temperature sensor comprises:

carrying out weighted average on the received signal and the latest N times of historical received signals, and taking an average result as the filtering signal; n is a positive integer.

7. The gesture recognition method of claim 5, wherein detecting the first trigger signal from the filtered signal comprises any one of:

when the deviation between the filtering signal and the adjacent latest M historical filtering signals is larger than a first deviation value, determining that the first triggering signal is detected, wherein M is a positive integer;

or determining that the first trigger signal is detected when the deviation between the filtered signal and the adjacent latest M historical filtered signals is larger than the first deviation value and the deviation between the filtered signal and other historical filtered signals is larger than the second deviation value; the other historical filtered signals are not adjacent to the last M historical filtered signals;

or, the deviation between the filtering signal and the X times of historical filtering signals is larger than the first deviation value, and the M times of historical filtering signals are in monotonically increasing relation, so that the first trigger signal is determined to be detected; the X times of historical filtering signals are selected from the M times of historical filtering signals;

or performing Fourier frequency domain transformation on the filtered signal and the M times of historical filtered signals, and determining that the first trigger signal is detected if a multi-order frequency offset component exists in a frequency domain transformation result.

8. The gesture recognition method of claim 7, further comprising:

determining the first deviation value and/or the second deviation value according to the temperature difference between the reference human body temperature and the ambient temperature; the first deviation value is positively correlated with the temperature difference; the second deviation value is positively correlated to the temperature difference.

9. The gesture recognition method of claim 1, further comprising:

detecting whether the optical window of the temperature sensor is polluted;

if the light window is detected to be polluted, generating and outputting reminding information; the reminding information is used for indicating that the light window is polluted.

10. The gesture recognition method of claim 9, wherein detecting whether the light window is contaminated comprises:

acquiring the current ambient temperature;

and if the difference between the current ambient temperature obtained by the temperature sensor and the ambient temperature obtained by the temperature sensor obtained by statistics is larger than a preset difference, determining that the optical window is polluted.

11. A gesture recognition apparatus, comprising:

the detection unit is used for detecting the first trigger signal;

the determining unit is used for determining whether possible gesture actions are found according to the first trigger signal; the first trigger signal is triggered by target induction area temperature jump induced by the temperature sensor.

12. A gesture recognition device, comprising: temperature sensor, amplifier circuit and microprocessor, wherein: the temperature sensor is suitable for detecting the temperature change of the target induction area, generating a temperature change signal and outputting the temperature change signal to the amplifying circuit;

the amplifying circuit is suitable for amplifying the temperature change signal and outputting the amplified temperature change signal to the microprocessor;

and the microprocessor is suitable for determining whether gesture actions exist according to the jump condition of the temperature change signal within the preset duration.

13. The gesture recognition device of claim 12, wherein the temperature sensor is a non-contact temperature sensor comprising any one of: an infrared thermopile sensor, a PIR infrared human body sensor and a vanadium oxide temperature sensor.

14. A computer readable storage medium, being a non-volatile storage medium or a non-transitory storage medium, having stored thereon a computer program, characterized in that the computer program when executed by a processor performs the steps of the gesture recognition method according to any of claims 1-10.