CN111083400A

CN111083400A - Pixel structure of double modulation frequency based on CPAD

Info

Publication number: CN111083400A
Application number: CN201811229490.7A
Authority: CN
Inventors: 徐江涛; 李嘉文; 苗津; 刘伯文; 张培文
Original assignee: Tianjin University Marine Technology Research Institute
Current assignee: Tianjin University Marine Technology Research Institute
Priority date: 2018-10-22
Filing date: 2018-10-22
Publication date: 2020-04-28

Abstract

A CPAD-based dual-modulation-frequency pixel structure is characterized in that each pixel comprises a dual FD node, the CPAD method is adopted to increase the charge transfer rate, so that the problem of poor imaging quality due to disordered phase information caused by too low charge transfer rate is solved, and the dual modulation frequency is adopted to ensure the precision of a three-dimensional image sensor under the condition of ensuring the detection distance.

Description

Pixel structure of double modulation frequency based on CPAD

Technical Field

The invention relates to the field of analog integrated circuit design, in particular to a CPAD (cross point display) -based dual-modulation-frequency pixel structure.

Background

Image sensors have been a focus of human research. However, with the development of modern science and technology, the requirement of human beings on the traditional 2D image sensor is higher and higher, and not only higher resolution, higher speed and larger dynamic range are expected, but also the human beings expect to obtain the depth information of the object, but the 2D imaging technology cannot meet the requirement of human beings at present, so the 3D imaging technology is developed accordingly. The TOF algorithm based on continuous waves is simple, high in measurement accuracy and more in use, and the main working principle is to detect the time difference between light emitted to an object and reflected to an image sensor so as to calculate the distance between the image sensor and the object.

At present, research on a TOF image sensor is greatly advanced, but still there are many technical difficulties, for example, in an indirect TOF sensor, near-infrared light received by a pixel is weak under the condition that an object is far away, which may cause the imaging accuracy of the sensor to be greatly reduced; meanwhile, when the charge transfer speed in the pixel is not high enough, the phenomenon of too low modulation contrast ratio can be caused, and the imaging of the sensor can be influenced.

Disclosure of Invention

Aiming at the problems in the prior art, the invention provides a CPAD (cross talk display device) -based dual-modulation-frequency pixel structure, each pixel comprises a dual FD (field-effect transistor) node, the CPAD method is adopted to increase the charge transfer rate, so that the problem of low imaging quality due to disordered phase information caused by too low charge transfer rate is solved, and the dual modulation frequency is adopted to ensure the precision of a three-dimensional image sensor under the condition of ensuring the detection distance.

A CPAD-based dual-modulation-frequency pixel structure is characterized in that a photoelectric detector consists of four electrodes in a CAPD pixel, and two electrodes are used for collecting signals (DET) by the detector_AAnd DET_B) Two for biasing to assist faster charge flow. V_AAnd V_BSwitching and modulating light sources at two different electrodes at the same frequency but opposite phase modulation frequency CLK_AAnd CLK_BTo respectively accelerate the transfer of collected photo-generated charge to different phases. After obtaining the electronic number representing the distance information of four different phases, the depth information of the illuminated object can be calculated by using a correlation formula, and the information of the four phases is collected so as to reduce the influence from the background light. In addition, it has some functional transistors, which have a charge transfer gate (TRG), a reset transistor (RST), a source follower Amplifier (AMP)_AAnd AMP_B) And the function of the pixel select transistor (SEL).

The charge transfer gate tube separates an N buried layer of the photosensitive photodiode from a silicon-silicon dioxide interface, so that the dark current of the pixel is greatly reduced, and related double sampling is realized by an FD node newly added relative to a 3T pixel, thereby greatly reducing random noise and inhibiting fixed mode noise. The reset transistor is used for resetting the FD node so as to collect a required exposure signal and a reset signal; the source follower is connected with the FD node, so that a stable signal is obtained; the pixel selection transistor can select a pixel signal to be read.

In addition, a dual modulation frequency is employed in this pixel, and an operation sequence is shown in fig. 2 (a). The lower frequency f1 (e.g., 80 MHz) acquires the first phase, and then the higher frequency f2 (e.g., 100 MHz) acquires the phase. Finally, a formula is applied that eliminates and combines the phases from frequencies f1 and f2 to produce the final depth result. The uncertainty (jitter) of the final calculated depth is about the same as the combined integration time using frequency (f 1 + f 2)/2 (e.g., 90 MHz), as shown in fig. 2 (b). It can be shown that the range over which explicit operation can be performed is many times greater than f1 or f2 (e.g., the operating range is typically proportional to f1-f2, in which case f1-f2 is 100MHz-80MHz =20 MHz).

A pixel structure based on CPAD double modulation frequency avoids the problem that the speed of four frames of two-dimensional image information to generate one frame of three-dimensional depth information is too slow, and simultaneously utilizes a current auxiliary photon demodulator to accelerate the charge transfer rate during exposure and avoid the problem of noise generated by unclean charge transfer to imaging; the double modulation frequency is used for detecting the approximate distance of the object at a lower modulation frequency, and then the accurate depth information of the object is obtained at a higher modulation frequency, so that the detection distance of the sensor is ensured, the accuracy of the sensor is improved on the basis of the distance, meanwhile, the exposure time is not increased, and the imaging problems such as blurring caused by overlong exposure time are avoided.

Drawings

Figure 1 is a diagram of a CPAD dual FD pixel architecture of the present invention;

fig. 2 is a sequence diagram of TOF multiple sequence operation.

Detailed Description

The invention will be further described with reference to the following figures and specific examples, but the scope of the invention is not limited thereto.

Whole pixel is because including two FD nodes, so when drawing the domain, this pixel area can be bigger a little than ordinary image sensor area, but is not more than 12um best, and the area is too big can lead to the parasitic effect problem to when the technology factory made, adopt the back of the body formula technology can have better photosensitive effect than the positive expression to under this technology, the modulation contrast of pixel increases and the rate of decay can reduce a lot along with the modulation frequency.

Claims

1. A pixel structure of double modulation frequency based on CPAD is characterized in that: the photodetector consists of four electrodes in one CAPD pixel, two for the detector to collect the signal and two for the bias to assist faster charge flow; v_AAnd V_BSwitching and modulating light sources at two different electrodes at the same frequency but opposite phase modulation frequency CLK_AAnd CLK_BTo accelerate the transfer of collected photo-generated charge of different phases, respectively; after the electronic number of the four different phases representing the distance information is obtained, the depth information of the illuminated object can be calculated by using a correlation formula, and the information of the four phases is collected so as to reduce the influence from background light; in addition, it has a charge transfer gate, a reset transistor, a source follower amplifier, and a pixel selection transistor.

2. The CPAD-based dual modulation frequency pixel structure of claim 1, wherein: the charge transfer gate tube separates the N buried layer of the photosensitive photodiode from the silicon-silicon dioxide interface, so that the dark current of the pixel is greatly reduced, and the related double sampling is realized by the newly added FD node relative to the 3T pixel, thereby greatly reducing the random noise and inhibiting the fixed mode noise.

3. The CPAD-based dual modulation frequency pixel structure of claim 1, wherein: the reset transistor is used to reset the FD node in order to acquire a desired exposure signal and reset signal.

4. The CPAD-based dual modulation frequency pixel structure of claim 1, wherein: the source follower is connected to the FD node, thereby obtaining a stable signal.

5. The CPAD-based dual modulation frequency pixel structure of claim 1, wherein: the pixel selection transistor can select a pixel signal to be read.