TWI533256B - Gesture sensing device and method of sensing three-dimensional gestures - Google Patents

Description

Image sensing device and image sensing method

The present invention relates to a 3D image sensing device and a 3D image sensing method, and more particularly to a 3D image sensing device and 3D image sensing method for eliminating background light noise by using an integer multiple of alternating current frequency or a predetermined frequency.

With the evolution of the human-machine interface of mobile devices, the human-machine interface has a tendency toward a natural user interface (NUI), in which gesture recognition is one of the most important natural user interface technologies. Gesture recognition can be performed using different 2D image processing. However, the 2D image sensor cannot determine the depth of the image, which greatly affects the ability of gesture recognition. Therefore, the use of 3D image sensing including depth information is a preferred method of performing gesture recognition in the future.

Ambient light sensors (ALS) and proximity sensors (PS) are widely used on most mobile devices (such as smart phones or tablets). Because for any mobile device, the space for new components placed on the printed circuit board will greatly increase the cost of the mobile device, any mobile device has an integrated background light sensor, adjacency sensor, color temperature sensor or temperature. The need for sensors and gestures to be recognized to the same integrated circuit. However, the prior art does not provide an integrated solution to the user.

An embodiment of the invention provides a 3D image sensing device. The 3D image sensing device comprises a light source, a sensing module and a signal processing module, wherein the sensing module comprises a pixel array, a control unit and a light source driver. The light source is for generating a scintillation light having a K times alternating current frequency or a predetermined frequency, wherein K is a positive real number. The pixel array is configured to sample the scintillation light to generate a sampling result, wherein the sampling result includes a plurality of first images when the light source is turned on and a plurality of second images when the light source is turned off; the control unit compares the sampling result Performing an image processing to generate a spectrum corresponding to the sampling result; the light source driver is configured to drive the light source according to the K times alternating current frequency or the predetermined frequency. The signal processing module is coupled to the sensing module for generating the K times alternating current frequency according to the frequency spectrum, and generating a depth information according to the plurality of first images and the plurality of second images, wherein The signal processing module outputs the K times alternating current frequency to the light source driver when the K times alternating current frequency is within a predetermined frequency range, or the signal processing module when the K times alternating current frequency is outside the predetermined frequency range The predetermined frequency is output to the light source driver.

Another embodiment of the present invention provides a 3D image sensing device, wherein the 3D image sensing device applied to the method includes a light source, a sensing module, and a signal processing module. The method includes the sensing module driving the light source to generate a flashing light having the K times alternating current frequency or the predetermined frequency according to a K times alternating current frequency or a predetermined frequency, wherein K is a positive real number; the sensing module Sampling the scintillation light to generate a sampling result, where the sampling result includes a plurality of first images when the light source is turned on and a plurality of second images when the light source is turned off; the sensing module performs an image on the sampling result Processing to generate a spectrum corresponding to the sampling result; the signal processing module directly outputs the predetermined frequency, or generates and outputs the K times alternating current frequency to the sensing module according to the spectrum; the signal processing module is based on the complex The first image and the plurality of second images generate a depth information.

The invention provides a 3D image sensing device and a method for 3D image sensing. Compared with the prior art, the 3D image sensing device integrates a sensing module and a signal processing module into the same integrated circuit, so the invention has lower cost and more efficiency. In addition, since a light source can generate a scintillation light having an integer multiple of an alternating current frequency or a predetermined frequency, the present invention can eliminate the noise of a background light, resulting in the ability of the depth sensing and gesture recognition of the present invention to be improved.

Please refer to FIG. 1 . FIG. 1 is a schematic diagram of a 3D image sensing device 100 according to an embodiment of the present invention. As shown in FIG. 1 , the 3D image sensing device 100 includes a light source 102 , a sensing module 104 , and a signal processing module 106 . The sensing module 104 includes a pixel array 1042 , a control unit 1044 , and a The light source driver 1046, and the control unit 1044 includes an analog-to-digital converter 10442 and a discrete Fourier transform (DFT) converter 10444. The light source 102 is an infrared light source or an infrared laser light source (for example, an infrared light emitting diode or an infrared laser diode), and is used to generate a K times alternating current frequency (for example, 100 Hz or 120 Hz) or a predetermined a flashing light of frequency, wherein the exposure time of the pixel array 1042 when the light source 102 is turned on is equal to the exposure time of the pixel array 1042 when the light source 102 is turned off, and K is a positive real number (eg, K is N or 1/N, where N is A positive integer greater than one). However, in another embodiment of the invention, light source 102 is a source of visible light. Pixel array 1042 is a G by H complementary metal-oxide-semiconductor (CMOS) image sensing pixel array, where G and H are positive integers. However, the present invention is not limited to the pixel array 1042 being a complementary MOS image sensing pixel array, that is, the pixel array 1042 may also be a Charge-coupled Device (CCD) image sensing pixel array. The pixel array 1042 includes an infrared filter, so the pixel array 1042 can attenuate other light waves than infrared rays. In an embodiment of the invention, the wavelength of the scintillation light produced by the light source 102 is 850 nm. When the light source 102 generates scintillation light having a K-fold alternating current frequency or a predetermined frequency, the pixel array 1042 of the sensing module 104 is configured to sample the scintillation light generated by the light source 102 to generate a sampling result SR, wherein the pixel array 1042 generates The sampling result SR includes a plurality of first images when the light source 102 is turned on and a plurality of second images when the light source 102 is turned off, and the light source 102 and the pixel array 1042 can be synchronized with the starting point of each picture. After the pixel array 1042 generates the sampling result SR, the control unit 1044 performs an image processing on the sampling result SR generated by the pixel array 1042 to generate a spectrum SP corresponding to the sampling result SR generated by the pixel array 1042, wherein the spectrum SP corresponds to The sampling result is a plurality of second images when the light source 102 is turned off in the SR. As shown in FIG. 1, the image processing performed by the control unit 1044 includes an analog digital converter 10442 that generates a digital signal DS based on the sampling result SR generated by the pixel array 1042, and a discrete Fourier converter 10444 according to the analog digital converter 10442. The generated digital signal DS produces a spectrum SP corresponding to the sampling result SR generated by the pixel array 1042. In addition, the light source driver 1046 is configured to generate a square wave to turn on or turn off the light source 102 according to the K times alternating current frequency or a predetermined frequency, wherein the square wave generated by the light source driver 1046 has a K times alternating current frequency or a predetermined frequency.

As shown in FIG. 1 , the signal processing module 106 is coupled to the pixel array 1042 and the control unit 1044 . Because the sampling result SR generated by the pixel array 1042 includes a plurality of first images when the light source 102 is turned on and a plurality of second images when the light source 102 is turned off, the signal processing module 106 can be based on the plurality of first when the light source 102 is turned on. The plurality of second images when the image and the light source 102 are turned off generate a depth information. That is, the signal processing module 106 subtracts a corresponding second image of the plurality of second images from each of the first images of the plurality of first images to generate depth information. Referring to FIG. 2, FIG. 2 is a schematic diagram illustrating the depth of information that the signal processing module 106 can generate without being affected by the background light when the light source 102 generates scintillation light having a K-fold alternating current frequency or a predetermined frequency. As shown in FIG. 2, when the scintillation light has a K-fold alternating current frequency or a predetermined frequency, the signal processing module 106 generates an area corresponding to the background light (for example, an indoor fluorescent lamp) according to the plurality of first images (the light source 102 is turned on). (Block I and Block III) will be equal to the area corresponding to the background light (Block II and Block IV) generated from the plurality of second images (light source 102 off). Therefore, as shown in the formula (1), the signal processing module 106 can eliminate the influence of the energy waveform of the indoor fluorescent lamp (100 Hz or 120 Hz) through the plurality of first images and the plurality of second images. As shown in the formula (1), the equation (1) holds at different initial phase angles θ and different exposure times X. In addition, the depth information generated by the signal processing module 106 is approximately inversely proportional to the square of the distance.

(1)

Where 0<X<π/2

Further, the formula (2) is a general formula corresponding to the formula (1):

(2)

As shown in equations (2) and 3, X1 is the integration time when the light source 102 is turned on, X2 is the integration time when the light source 102 is off, and θ is the initial phase angle, where K can be N or 1/N, but N is a positive integer not equal to 1, and K*(X1+X2) is not greater than 1/120 Hz or 1/100 Hz.

As shown in FIG. 1, the signal processing module 106 includes a flicker noise discriminator 1062 and a frequency generator 1064. The frequency generator 1064 is a delay time oscillator and a RC oscillator. RC oscillator), a phase-locked loop, a crystal oscillator, a resonator, or a frequency divider. The flicker noise discriminator 1062 is configured to determine a flicker noise corresponding to the K times alternating current frequency according to the frequency spectrum SP, and to generate a flash corresponding to the K times alternating current frequency according to the flicker noise corresponding to the K times alternating current frequency. The frequency generator 1064 is coupled to the scintillation noise discriminator 1062 for generating a K-fold alternating current frequency AF to the light source driver 1046 according to the control signal CS, wherein the K-times alternating current frequency generated by the frequency generator 1064 is Within a predetermined frequency range. In addition, when the K-fold alternating current frequency generated by the frequency generator 1064 is outside the predetermined frequency range, the frequency generator 1064 directly outputs the predetermined frequency to the light source driver 1046.

Please refer to FIG. 4 . FIG. 4 is a schematic diagram illustrating the signal processing module 106 determining the gesture of a user by using the K-frequency alternating current frequency or a predetermined frequency generated by the light source 102. As shown in FIG. 2, when the light source 102 generates scintillation light having a K-fold alternating current frequency or a predetermined frequency, the signal processing module 106 can subtract each first image of the plurality of first images generated by the pixel array 1042. A corresponding second image of the plurality of second images generates depth information. Then, the signal processing module 106 can further determine the gesture of the user from the depth information.

Please refer to FIG. 5. FIG. 5 is a schematic diagram of a 3D image sensing device 400 according to another embodiment of the present invention. As shown in FIG. 5 , the difference between the 3D image sensing device 400 and the 3D image sensing device 100 is that the sensing module 404 of the 3D image sensing device 400 further includes a background visible light sensor 4042 and a proximity. A sensor 4044, a color temperature sensor 4046, and a temperature sensor 4048. Therefore, the sensing module 404 integrating the background visible light sensor 4042, the adjacent sensor 4044, the color temperature sensor 4046, and the temperature sensor 4048 has fewer component counts, occupies less board space, and is lower. the cost of. In addition, since the sensing module 404 integrates the background visible light sensor 4042, the adjacent sensor 4044, the color temperature sensor 4046, and the temperature sensor 4048, the sense of depth of the signal processing module 106 of the 3D image sensing device 400 The ability to measure and recognize gestures can be improved.

When the background visible light is sufficient and the light source 102 is turned off, the sampling result SR produced by the pixel array 1042 can still produce an outline of an object by the background visible light. Conversely, when the background visible light is dimmed, at this point, the light source 102 must be turned on to cause the pixel array 1042 to generate a sampling result SR for depth sensing and gesture recognition. Therefore, the background visible light sensor can dynamically adjust the duty cycle of the light source 102. As such, the 3D image sensing device 400 can further reduce the turn-on time of the light source 102 and optimize the overall power consumption of the depth sensing and gesture recognition of the 3D image sensing device 400. However, in most cases, the signal processing module 106 generates a depth information according to the plurality of first images when the light source 102 is turned on and the plurality of second images when the light source 102 is turned off, wherein the depth information is used for the signal processing mode. Group 106 identifies objects and backgrounds that are quite useful.

The adjacency sensor 4044 is based on the opening and closing of the light source 102, and detects the average distance between the object and the adjacent sensor 4044. Because the adjacency sensor 4044 can be used to detect the approximate distance between the object and the adjacent sensor 4044, the adjacency sensor 4044 can reduce the overall power consumption of the 3D image sensing device 400 as a whole and the reaction time is faster.

In an embodiment of the invention, the blocking infrared and color filter coatings are included in the color temperature sensor 4046, wherein the color temperature sensor 4046 can be measured by a spectrum corresponding to the blocking infrared and color filter coatings. Background light. In general, blocking infrared and color filter coatings can block infrared light having a wavelength of about 700 nm. In addition, the color temperature sensor 4046 can detect the color temperature of the background visible light, wherein the color temperature of the background visible light can be used for dynamically adjusting the white balance of a display device and enhancing the image quality of an image capturing device.

Temperature sensor 4048 is used to measure the temperature of the surrounding environment. However, measuring the temperature of the surrounding environment is also used to compensate for the temperature of the background light and color measurements. As shown in FIG. 5, the signal processing module 106 has a non-volatile memory (eg, a flash memory) 1066. The signal processing module 106 can perform gesture recognition based on the difference between the features extracted from a current sampling result SR and the pre-recorded features stored in the non-volatile memory 1066. When the difference is small enough, the signal processing module 106 can obtain a positive recognition.

Please refer to FIG. 6. FIG. 6 is a schematic diagram of a 3D image sensing device 500 according to another embodiment of the present invention. As shown in FIG. 6 , the difference between the 3D image sensing device 500 and the 3D image sensing device 400 is that the signal processing module 506 of the 3D image sensing device 500 further includes a touch controller 5062 . The benefit of integrating the gesture recognition and touch controller 5062 is that it has a lower overall circuit cost.

In addition, the 3D image sensing device 500 has the following advantages: First, the touch controller 5062 is mainly used to process a signal TS generated from the touch sensor 508, wherein the signal TS generated from the touch sensor 508 can be Represents the position of an object (such as a finger) that touches a touch panel or the position of the touch panel adjacent to the object (without touching the touch panel). Pixel array 1042 can be used to generate an indication direction of the object. Therefore, regardless of whether the object touches or moves away from the touch panel, combining touch and gesture recognition enables a user to have the ability to track an item on the display panel. Secondly, in an embodiment of the invention, when a user's finger is touching an item on the display panel, the user's other hand can perform a 3D gesture to operate the touched item on the display panel. . For example, a building in a map displayed by the display panel can be confirmed by the user touching, and then the 3D gesture of the user's other hand can perform the user's desire to move relative to the building in the map. Direction (such as up, down, left, right; zoom in or out) or an instruction in the direction of the building in the map.

Please refer to FIG. 7. FIG. 7 is a schematic diagram of a 3D image sensing device 600 according to another embodiment of the present invention. As shown in FIG. 7 , the difference between the 3D image sensing device 600 and the 3D image sensing device 400 is that the signal processing module 606 of the 3D image sensing device 600 further includes a digital image processor 6062. As shown in FIG. 7, an external image sensor 608 can be used to capture a 2D red, green, and blue image IS of an object. The digital image processor 6062 includes some normal image processing functions such as image compression and anti-shake.

In addition, the 3D image sensing device 600 has the following advantages: First, the signal processing module 606 can perform gesture recognition using not only the depth information but also the 2D red, green, and blue images IS. Because the red, green, and blue image sensors of the external image sensor 608 have a higher resolution, the signal processing module 606 can more effectively distinguish an object from a foreground. In addition, the overlap of the depth information and the 2D red, green and blue image IS can greatly enhance the accuracy of the extraction features. Second, when the user is in a particular situation (eg, the user wears a black finger cuff), the black finger cuff will weaken the infrared light reflected by the finger. However, the red, green, and blue image sensors of the external image sensor 608 have the ability to distinguish black finger sleeves from the background garment. Therefore, the complementary effect of the red-green-blue image sensor of the external image sensor 608 and the infrared sensor in the pixel array 1042 can enhance the accuracy of the gesture recognition capability of the 3D image sensing device 600.

Please refer to FIG. 8. FIG. 8 is a schematic diagram of a 3D image sensing device 700 according to another embodiment of the present invention. As shown in FIG. 8 , the difference between the 3D image sensing device 700 and the 3D image sensing device 400 is that the signal processing module 706 of the 3D image sensing device 700 includes a sensor hub of an inertial sensor 708 . 7062, wherein the inertial sensor 708 comprises a Micro Electro Mechanical Systems (MEMS) accelerometer, a gyroscope, a magnetometer, and an altimeter. The inertial sensor 708 is used as an inertial measurement unit (IMU), wherein the inertial measurement unit is quite important for indoor navigation. Because the signal processing module 706 integrates the sensing hub 7062 of the inertial sensor 708 and the gesture processor, the signal processing module 706 is a more efficient integrated circuit.

In addition, the 3D image sensing device 700 has the following advantages: First, the sensing hub 7062 of the inertial sensor 708 can distinguish the position, direction, and tilt of the mobile phone user. The sensing module 404 includes a background visible light sensor, an adjacent sensor, a color temperature sensor, and a temperature sensor. Thus, the location and surroundings of the user can be recognized by the 3D image sensing device 700, and many applications can be generated depending on the location of the user and the surrounding environment. Second, because the user can move the mobile phone around itself, the measured direction of the mobile phone can be performed by the gyroscope and the magnetometer. The signal processing module 706 can then reconstruct the 3D image of the user. In another embodiment of the invention, the measurement of the tilt and/or direction of the inertial measurement unit can assist the mobile device in detecting changes in the viewing direction. Because the mobile device may leave the detection angle or detection range, the inertial measurement unit (inertial sensor 708) and the gesture processor will alert the user to the gesture or detection range of the pixel array 1042.

In addition, the 3D image sensing devices 100, 400, 500, 600, and 700 are not limited to determining only the gesture of the user, that is, the 3D image sensing devices 100, 400, 500, 600, and 700 can also identify the user. Finger, back of the hand or eyeball, or perform object recognition.

Referring to FIG. 1 , FIG. 2 , FIG. 4 , and FIG. 9 , FIG. 9 is a flowchart illustrating a method for 3D image sensing according to another embodiment of the present invention, wherein the method of FIG. 9 utilizes The 3D image sensing device 100 of FIG. 1 illustrates the detailed steps as follows:

Step 800: Start;

Step 802: The sensing module 104 drives the light source 102 to generate a flashing light having a K times alternating current frequency or a predetermined frequency according to a K times alternating current frequency or a predetermined frequency;

Step 804: The sensing module 104 samples the scintillation light to generate a sampling result SR, where the sampling result SR includes a plurality of first images when the light source 102 is turned on and a plurality of second images when the light source 102 is turned off;

Step 806: The analog-to-digital converter 10442 generates a digital signal DS according to the sampling result SR;

Step 808: The discrete Fourier converter 10444 generates a spectrum SP corresponding to the sampling result SR according to the digital signal DS;

Step 810: The blinking noise discriminator 1062 determines a blinking noise corresponding to the K times alternating current frequency according to the frequency spectrum SP.

Step 812: The flicker noise discriminator 1062 generates a control signal CS corresponding to the K times AC frequency according to the flicker noise corresponding to the K times AC frequency;

Step 814: The frequency generator 1064 directly outputs a predetermined frequency, or generates a K times alternating current frequency AF to the sensing module 104 according to the control signal CS;

Step 816: The signal processing module 106 generates a depth information according to the plurality of first images and the plurality of second images.

In step 802, the light source driver 1046 in the sensing module 104 is configured to generate a square wave to turn on or off the light source 102 according to the K times alternating current frequency or a predetermined frequency, wherein the square wave generated by the light source driver 1046 has K times alternating current Frequency or predetermined frequency. In addition, the light source 102 is an infrared light source or an infrared laser light source (for example, an infrared light emitting diode or an infrared laser diode), and is used to generate a K times alternating current frequency (for example, 100 Hz or 120 Hz) or predetermined. Frequency flickering light, wherein the exposure time of the pixel array 1042 when the light source 102 is turned on is equal to the exposure time of the pixel array 1042 when the light source 102 is turned off, and K is a positive real number (eg, K is N or 1/N, where N is one A positive integer greater than one). In step 804, when the light source 102 generates scintillation light having a K-fold alternating current frequency or a predetermined frequency, the pixel array 1042 of the sensing module 104 is used to sample the scintillation light generated by the light source 102 to generate a sampling result SR, wherein the light source 102 and pixel array 1042 can be synchronized to the beginning of each picture. In step 806, after the pixel array 1042 generates the sampling result SR, the analog digital converter 10442 in the control unit 1044 generates the digital signal DS according to the sampling result SR generated by the pixel array 1042, and in step 808, the control unit 1044 The discrete Fourier transformer 10444 generates a spectrum SP corresponding to the sampling result SR generated by the pixel array 1042 according to the digital signal DS generated by the analog-to-digital converter 10442, wherein the spectrum SP corresponds to the sampling result SR when the light source 102 is turned off. A plurality of second images.

In step 810 and step 812, the flicker noise discriminator 1062 in the signal processing module 106 is configured to determine flicker noise corresponding to the K times alternating current frequency according to the frequency spectrum SP, and to use the frequency corresponding to the K times alternating current frequency. The flicker noise produces a control signal CS corresponding to the K times the AC frequency. In step 814, the frequency generator 1064 in the signal processing module 106 generates a K times alternating current frequency AF to the light source driver 1046 in the sensing module 104 according to the control signal CS, wherein the frequency generator 1064 generates K times alternating current The frequency is within a predetermined frequency range. In addition, when the K-fold alternating current frequency generated by the frequency generator 1064 is outside the predetermined frequency range, the frequency generator 1064 directly outputs the predetermined frequency to the light source driver 1046.

In step 816, as shown in FIG. 1 , the sampling result SR generated by the pixel array 1042 includes a plurality of first images when the light source 102 is turned on and a plurality of second images when the light source 102 is turned off, so the signal processing module 106 may generate depth information according to the plurality of first images when the light source 102 is turned on and the plurality of second images when the light source 102 is turned off. That is, the signal processing module 106 subtracts a corresponding second image of the plurality of second images from each of the first images of the plurality of first images to generate depth information.

In addition, as shown in FIG. 4, when the light source 102 generates scintillation light having a K-fold alternating current frequency or a predetermined frequency, the signal processing module 106 can utilize each of the plurality of first images generated by the pixel array 1042. Subtracting a corresponding second image of the plurality of second images to generate depth information. Then, the signal processing module 106 can further determine the gesture of the user from the depth information.

In summary, the present invention has lower cost and more efficiency than the prior art, because the 3D image sensing device provided by the present invention integrates the sensing module and the signal processing module to the same integrated circuit. . In addition, since the light source can generate scintillation light having an integer multiple of the alternating current frequency or a predetermined frequency, the present invention can eliminate the noise of the background light, resulting in the ability of the depth sensing and gesture recognition of the present invention can be improved.

The above are only the preferred embodiments of the present invention, and all changes and modifications made to the scope of the present invention should be within the scope of the present invention.

100, 400, 500, 600, 700 3D image sensing device 102 light source 104, 404 sensing module 106, 506, 606, 706 signal processing module 1042 pixel array 1044 control unit 1046 light source driver 10442 Analog to Digital Converter 10444 Discrete Fourier Transformer 1062 Scintillation Noise Discriminator 1064 Frequency Generator 1066 Non-volatile Memory 4042 Background Visible Light Sensor 4044 Adjacent Sensor 4046 Color Temperature Sensor 4048 Temperature Sensor 5062 Touch Controller 508 Touch Sensor 6062 Digital Image Processor 608 External Image Sensor 708 Inertial Sensor 7062 Hub AF sensing AC frequency control signal DS CS digital RGB image signal IS SP sampling result SR TS Spectrum Signal steps 800-816

Fig. 1 is a schematic view showing a 3D image sensing device according to an embodiment of the present invention. 2 and 3 are diagrams illustrating that the signal processing module can generate depth information that is not affected by the background light when the light source generates scintillation light having a K-fold alternating current frequency or a predetermined frequency. Fig. 4 is a schematic diagram for explaining the gesture of the user by the signal processing module using the scintillation light having a K-fold alternating current frequency generated by the light source. Fig. 5 is a schematic view showing a 3D image sensing device according to another embodiment of the present invention. Figure 6 is a schematic view showing a 3D image sensing device according to another embodiment of the present invention. Fig. 7 is a schematic view showing a 3D image sensing device according to another embodiment of the present invention. Figure 8 is a schematic view showing a 3D image sensing device according to another embodiment of the present invention. Figure 9 is a flow chart illustrating a method of 3D image sensing in accordance with another embodiment of the present invention.

100 3D image sensing device 102 light source 104 sensing module 106 signal processing module 1042 pixel array 1044 control unit 1046 Light source driver analog to digital converter 10442 10444 1062 discrete Fourier transformer 1064 scintillator noise discriminating AF oscillator frequency alternating current control signal CS DS Digital signal sampling result SR SP spectrum

Claims (27)

An image sensing device includes: a light source for generating a scintillation light; and a sensing module comprising: a pixel array for sampling the scintillation light to generate a sampling result, wherein the sampling result includes the light source being turned on a plurality of first images and a plurality of second images when the light source is turned off; a control unit performing image processing on the sampled results to generate a spectrum corresponding to the sampled result; wherein the spectrum corresponds to the sampling result a plurality of second images when the light source is turned off; and a light source driver connected to the light source for driving the light source; and a signal processing module coupled to the sensing module for generating a frequency according to the spectrum And driving the light source through the light source driver to generate scintillation light having a flicker frequency corresponding to the frequency to eliminate background noise. The image sensing device of claim 1, wherein the flickering light has a blinking frequency of a K times frequency or a predetermined frequency, wherein K is a positive real number. The image sensing device of claim 1 or 2, wherein the signal processing module generates a depth information based on the plurality of first images and the plurality of second images. The image sensing device of claim 1, wherein the light source is an infrared light source or an infrared laser light source. The image sensing device of claim 1, wherein the control unit comprises: an analog-to-digital converter for generating a digital signal according to the sampling result; and a discrete Fourier transform (DFT) converter, And according to the digital signal, the spectrum corresponding to the sampling result is generated. The image sensing device of claim 1 or 2, wherein the signal processing module comprises: a blinking noise discriminator for determining a blinking noise corresponding to the frequency according to the frequency spectrum, and for The flickering noise generates a control signal corresponding to the frequency; and a frequency generator coupled to the scintillation noise discriminator for generating the frequency according to the control signal. The image sensing device of claim 6, wherein the frequency generator is a delay time oscillator, a RC oscillator, a phase locked loop, a crystal oscillator, and a Resonator or a frequency divider. The image sensing device of claim 1, further comprising a touch sensor for generating a signal representing a position of the object touching the touch panel or a position of the touch panel adjacent to the object; and The signal processing module includes: a non-volatile memory; and a touch controller for processing the signal generated by the touch sensor. The image sensing device of claim 1, further comprising an external image sensor for capturing a 2D red, green and blue image of an object; and the signal processing module comprises: a non-volatile memory; And a digital image processor for performing image processing on the 2D red, green and blue image of the object. The image sensing device of claim 1 further comprising an inertial sensor, which is an inertial measurement unit; and the signal processing module comprises: a non-volatile memory; and a sensing hub for The position, direction and tilt of the image sensing device are discriminated based on the information of the inertial sensor. The image sensing device of claim 1, wherein the sensing module comprises a background visible light sensor, so that the working period of the light source is dynamically adjusted. The image sensing device of claim 1, wherein the sensing module comprises an adjacency sensor for detecting a distance between the object and the adjacent sensor according to the opening and closing of the light source. The image sensing device of claim 1, wherein the sensing module comprises a color temperature sensor for measuring background light. The image sensing device of claim 1, wherein the sensing module comprises a temperature sensor for measuring the temperature of the surrounding environment, wherein the temperature of the surrounding environment is used for the temperature of the background light and the color measurement. make up. The image sensing device according to any one of claims 1 to 8, wherein the signal processing module is further configured to: according to a first image of the plurality of first images and a corresponding one of the first images The second image determines a user's gesture. A method for image sensing, wherein the image sensing device for the method comprises a light source, a sensing module and a signal processing module, the method comprising: the sensing module is configured to drive the light source to generate a flashing The sensing module samples the scintillation light to generate a sampling result, where the sampling result includes a plurality of first images when the light source is turned on and a plurality of second images when the light source is turned off; the sensing module Performing an image processing on the sampling result to generate a spectrum corresponding to the sampling result; wherein the spectrum corresponds to a plurality of second images when the light source is turned off in the sampling result; and the signal processing module generates and outputs according to the spectrum a frequency, and the light source is driven according to the frequency by the sensing module, so that the blinking frequency of the flashing light generated by the light source corresponds to a frequency to eliminate background noise. The method of claim 16, wherein the scintillation light has a blinking frequency of a K times frequency or a predetermined frequency, wherein K is a positive real number, and K is N or 1/N, wherein N is a positive integer greater than one. . The method of claim 16 or 17, further comprising: the signal processing module generating a depth information according to the plurality of first images and the plurality of second images. The method of claim 17, the signal processing module directly outputs a predetermined frequency or generates and outputs the K times frequency to the sensing module according to the spectrum, and the sensing module is based on the predetermined frequency or the K The multiple frequency, the blinking frequency of the flashing light that drives the light source is the K times frequency or the predetermined frequency. The method of claim 16, wherein the exposure time of the pixel array when the light source is turned on is equal to the exposure time of the pixel array when the light source is turned off. The image sensing device of the method of claim 16, further comprising a touch sensor, wherein the sensing group comprises an array of pixels, the signal processing module comprising a non-volatile memory And a touch controller, wherein the method further includes: the touch controller is configured to process a signal generated from the touch sensor, wherein the signal generated from the touch sensor represents a touch panel The position of the object or the position of the touch panel adjacent to the object; the pixel array is used to generate an indication direction of the object. The method of claim 16, wherein the image sensing device further comprises an external image sensor, and the signal processing module comprises a non-volatile memory and a digital image processor; The method further includes: the external image sensor is configured to capture a 2D red, green, and blue image of an object, and the signal processing module performs gesture recognition using the depth information or the 2D red, green, and blue image. The method of claim 16, wherein the image sensing device applied to the method further comprises an inertial sensor, and the signal processing module comprises a non-volatile memory and a sensing hub; wherein the method further The method includes: the inertial sensor is an inertial measurement unit, and the sensing hub distinguishes the position, direction, and inclination of the image sensing device according to the information of the inertial sensor. The method of claim 16, wherein the sensing module of the image sensing device of the method comprises a background visible light sensor; wherein the method further comprises: the background visible light sensor for dynamically adjusting the light source Work cycle. The method of claim 16, wherein the sensing module of the image sensing device applied to the method comprises an adjacency sensor; wherein the method further comprises: the adjacency sensor for turning on and off according to the light source , detecting the average distance between the object and the adjacent sensor. The method of claim 16, wherein the sensing module of the image sensing device of the method comprises a color temperature sensor; wherein the method further comprises: the color temperature sensor for measuring the background light. The method of claim 16, wherein the sensing module of the image sensing device of the method comprises a temperature sensor; wherein the method further comprises: the temperature sensor for measuring the temperature of the surrounding environment The temperature of the surrounding environment is used to compensate for the temperature of the background light and color measurement.