CN102261958B - Method for reading out high-accuracy signal - Google Patents

Abstract

The invention discloses a method for reading out a high-accuracy signal in the technical field of micro-electrons and photoelectrons. The method is applicable to the field of an array type infrared image sensor. The array type infrared image sensor is mainly provided with an infrared detector array and a column-level read-out circuit; dark-pixel references are arranged around the conventional pixel array; firstly, the data of an upper reference row and a lower reference row is stored, and row compensation values of the data of all rows are calculated; secondly, pixels of all rows in the infrared detector array are sequentially gated row by row to finish reading out of frame data, and in the method, row compensation values and column compensation values of all columns of data in the current row are calculated by a calibration circuit, and the data of the infrared detector array is subjected to two-dimensional linear compensation; and finally, column gating logic is generated by a column selection circuit, so all columns of the compensated data of the current row are read out of a chip. By the method, background signals of pixels can be effectively inhibited, the output accuracy of a read-out circuit is improved, and the problem of non-consistence caused by relative deviation of a process can be eliminated by a simpler circuit.

Description

A kind of high-precision signal reading method

Technical field

The invention belongs to microelectronics and photoelectron technology field, relate in particular to a kind of high-precision signal reading method.

Background technology

Infrared imagery technique is a kind ofly to convert sightless infrared radiation signal the technology of visual picture into, and it all is widely used for example night vision, medical treatment, supervision etc. at military, civil area.Infrared image sensor comprises infrared detector array and sensing circuit two parts as the core of infrared imagery technique.Type according to detector can be divided into refrigeration mode and non-refrigeration type with infrared image sensor.Wherein non-refrigeration type has light, advantage cheaply, in recent years, along with the development of non-refrigeration type detector technology, has become main flow at civil area.The major function of sensing circuit is that the output signal to infrared eye amplifies, and analog to digital conversion and reading provides various low noise offset signals for infrared eye simultaneously.Common sensing circuit mainly selects circuit to form by row reading circuit unit, timing sequence generating circuit and ranks.

The principle of work of non-refrigerated infrared detector is: the energy of infrared radiation is converted into heat on the absorbing structure of pixel, thereby has changed the temperature of pixel, and detector is an electric signal with temperature inversion again.Because the ambient temperature of pixel is also read simultaneously, so it not only influences the dynamic range of sensing circuit, and its fluctuation also can be carried in the read output signal.Usually the non-refrigeration type infrared image sensor need be operated among the thermoelectric (al) cooler TEC of temperature constant, thereby has produced bigger system power dissipation.In recent years, the method that adopts dark pixel to eliminate background signal once was suggested, and this method cuts background signal through not receiving infrared radiation dark pixel detection background signal again from read output signal, thereby had suppressed the influence of background signal.Yet in the sensor for large scale array, background signal is not in full accord, and therefore this method has than mistake.Along with the focal plane Pixel Dimensions is more and more littler, the array scale is increasing in addition, and the fixed pattern noise FPN of the focal plane that the technology relative deviation (nonuniformity) of pixel is caused will be more and more obvious.Problem hereto adopts the back Calibration Method usually, promptly after completion is made in the focal plane, under radiationless condition, carry out a frame and read, and the relative deviation of each pixel is recorded in the storer, thereby the read output signal during to operate as normal compensates.This method has not only increased testing cost, and has increased fairly large storer and complicated hardware such as calibration circuit as being used to write down, has therefore increased big cost.

Summary of the invention

Deficiencies such as the existing method system power dissipation to mentioning in the above-mentioned background technology is higher, hardware size is bigger the present invention proposes a kind of high-precision signal reading method.

Technical scheme of the present invention is; A kind of high-precision signal reading method; Be applied to the array infrared image sensor; This sensor comprises infrared detector array, upward reference line, following reference line, left reference columns, right reference columns, row are selected circuit, row reading circuit, column select circuit and calibration circuit, it is characterized in that this method may further comprise the steps:

Step 1: row is selected circuit reference line and following reference line on the gating successively, and last reference line and following reference line are preserved output data behind row reading circuit;

Step 2: set i=1;

Step 3: row is selected the capable pixel of circuit gating i, obtains the data of this row through row reading circuit, and calibration circuit calculates the capable offset of this line data through the output data of row reading circuit in the step 1 simultaneously; Calibration circuit calculates the row offset of this each column data of row through left reference columns and right reference columns;

Step 4: capable each column data of i is carried out the two-dimensional linear compensation through its corresponding capable offset and row offset, and the data after the compensation that this row respectively is listed as through column select circuit are again read successively;

Step 5: make i=i+1, repeating step 3 and step 4 are read until a frame pixel.

The computing formula of said capable offset is:

$R_{i,j}=\frac{i}{2N}{D}_{\mathrm{up},j}+\frac{N-i}{2N}{D}_{\mathrm{down},j}$

Wherein:

R _{I, j}The capable offset of representing the capable j row of i pixel;

N is the line number of detector array;

D _UpBe last reference line;

D _DownBe following reference line;

D _{Up, j}Be D _UpThe j column data;

D _{Down, j}Be D _DownThe j column data.

The computing formula of said row offset is:

$C_{i,j}=\frac{j}{2M}{D}_{\mathrm{left},i}+\frac{M-j}{2M}{D}_{\mathrm{right},i}$

Wherein:

C _{I, j}The row offset of representing the capable j row of i pixel;

M is the columns of detector array;

D _LeftBe left reference columns;

D _RightBe right reference columns;

D _{Left, i}Be D _LeftThe i line data;

D _{Right, i}Be D _RightThe i line data.

The inventive method is carried out the two-dimensional linear compensation to the sensing circuit of infrared image sensor, can suppress the background signal of pixel effectively, improves the output accuracy of sensing circuit.And the nonuniformity problem that can be brought with better simply circuit for eliminating technology relative deviation.In the calibration circuit of the inventive method, through hardware usefulness technology, every row can only adopt one simply to take advantage of and add circuit, have taken less chip area, are fit to extensive infrared image sensor array.

Description of drawings

Fig. 1 is an infrared image sensor sensing circuit embodiment synoptic diagram of the present invention;

Fig. 2 is the calibration circuit process flow diagram of sensing circuit embodiment of the present invention.

Embodiment

Below in conjunction with accompanying drawing, preferred embodiment is elaborated.Should be emphasized that following explanation only is exemplary, rather than in order to limit scope of the present invention and application thereof.

Of the present inventionly proposed a kind of high-precision signal reading method that is used for infrared image sensor, utilized to be looped around the outer circle dark pixel of infrared detector array signal as a reference, the read output signal of infrared detector array has been carried out the two-dimensional linear compensation.

Key step of the present invention is:

A kind of high-precision signal reading method; Be applied to the array infrared image sensor; This sensor comprises infrared detector array, upward reference line, following reference line, left reference columns, right reference columns, row are selected circuit, row reading circuit, column select circuit and calibration circuit, it is characterized in that this method may further comprise the steps:

1: row is selected circuit reference line and following reference line on the gating successively, and last reference line and following reference line are preserved output data behind row reading circuit;

2: set i=1;

3: row is selected the capable pixel of circuit gating i, obtains the data of this row through row reading circuit, and calibration circuit calculates the capable offset of this line data through the output data of row reading circuit in 1 simultaneously; Calibration circuit calculates the row offset of this each column data of row through left reference columns and right reference columns;

4: capable each column data of i is carried out the two-dimensional linear compensation through its corresponding capable offset and row offset, and the data after the compensation that this row respectively is listed as through column select circuit are again read successively;

5: make i=i+1, repeat 3 and 4, read until a frame pixel.

Infrared image sensor is provided with infrared detector array, upward reference line (going up the dark pixel reference line), following reference line (following dark pixel reference line), left reference columns (left dark pixel reference columns), right reference columns (right dark pixel reference columns), row reading circuit, calibration circuit, row are selected circuit, column select circuit.

Last reference line, down reference line, left reference columns and right reference columns lay respectively at upper and lower a, left side and the right side of infrared detector array, are used to produce two-dimensional reference signal.

Row reading circuit can be realized amplifying and analog-digital conversion function, and it amplifies the picture element signal tiny signal, and converts digital signal into, is used for calibration circuit and handles.

Row selects the logic of circuit to be: gating two row reference pixels up and down at first respectively, and each row pixel in the gating infrared detector array line by line successively again, calibration circuit can at first calculate the trip calibration value and each is gone sense data compensates like this.

Calibration circuit carries out the two-dimensional linear compensation through ranks dark pixel signal data to the data of detector pixel, is used for eliminating background signal and process deviation.

Calibration circuit combines row to select circuit to have following logic: storage is the reference line data up and down; Calculate the capable offset of current line; Calculate the row offset of each row of current line, wherein the computing method of ranks offset are linear interpolation; Compensation current line pixel data, promptly each column data of current line subtracts each other with corresponding ranks offset; In conjunction with column select circuit each row compensation back data of current line are read out to outside the sheet.

Fig. 1 is an infrared image sensor sensing circuit embodiment synoptic diagram of the present invention, and is as shown in the figure, and the sensing circuit of present embodiment comprises the detector array of M * N scale, is used to survey infrared radiation and converts electric signal into; Be looped around two row, the two row dark pixels of detector one circle, be respectively left reference columns, go up reference line, right reference columns and following reference line, be used to sensing circuit two-dimensional reference signal is provided, so that the calibration circuit offset value calculation compensates each pixel; Row is selected circuit, is used to produce row and selects logic, so that row reading circuit is read a certain row; Row reading circuit, each row sensing element comprises amplifying circuit, analog to digital conversion circuit, be used for to pixel output signal amplify, analog to digital conversion, the digital signal that obtains is easy to computing; Calibration circuit through two row reference signals are calculated capable offset, calculates the row offset through two row reference signals, thereby can carry out the two-dimensional linear compensation to sense data; Column select circuit produces the logical logic of column selection, thereby data after the compensation of one-row pixels is read out to outside the sheet.

In the row sensing element of row reading circuit, amplifying circuit can be selected amplifying circuit commonly used for use according to type photodetector, for example source electrode follower type SFD, directly injection type DI or electric capacity transimpedance amplifier type CTIA; Analog to digital converter can be selected circular form analog to digital conversion circuit or integral form analog to digital conversion circuit according to the consideration of speed and precision.

Calibration circuit can adopt numeral to take advantage of to add circuit to realize that its operating process is as shown in Figure 2.Storage is the reference line sense data up and down, when promptly each frame is read beginning, at first distinguishes gating two row dark pixels up and down, obtains the digital sense data D of reference line through row reading circuit _UpDigital sense data D with following reference line _Down, and be stored in the register of calibration circuit;

Calculate the capable offset of current line, promptly through the row of two in register reference data D _UpAnd D _Down, the capable offset of calculating current line, computing formula is following:

$R_{i,j}=\frac{i}{2N}{D}_{\mathrm{up},j}+\frac{N-i}{2N}{D}_{\mathrm{down},j}$

Wherein:

R _{I, j}The capable offset of representing the capable j row of i pixel;

N is the line number of detector array;

D _{Up, j}Be D _UpThe j column data;

D _{Down, j}Be D _DownThe j column data.

M row row compensation value calculation can be accomplished through M multiplicaton addition unit parallel work-flow, has obtained M capable capable offset of i.

Calculate the row offset of each row of current line, the capable left side of the i reference columns data D that promptly reads through row reading circuit _Left, right reference columns data D _RightCalculate each row offset, computing formula is following:

$C_{i,j}=\frac{j}{2M}{D}_{\mathrm{left},i}+\frac{M-j}{2M}{D}_{\mathrm{right},i}$

Wherein:

C _{I, j}The row offset of representing the capable j row of i pixel;

M is the columns of detector array;

D _{Left, i}Be D _LeftThe i line data;

D _{Right, i}Be D _RightThe i line data.

Above-mentioned M multiplicaton addition unit parallel work-flow of M row row compensation value calculation reusable accomplished, and obtained M capable row offset of i.

Each row sense data is subtracted each other with corresponding ranks offset, the ranks offset R that the capable M row of i sense data is corresponding with these row _{I, j}And C _{I, j}Subtract each other, obtain final sense data, and export to outside the sheet successively through column select circuit control.Repeat above step the i+1 line data is read, all read completion until the N line data.

The above; Be merely the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, any technician who is familiar with the present technique field is in the technical scope that the present invention discloses; The variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.

Claims (1)

1. high-precision signal reading method; Be applied to the array infrared image sensor; This sensor comprises infrared detector array, upward reference line, following reference line, left reference columns, right reference columns, row are selected circuit, row reading circuit, column select circuit and calibration circuit, it is characterized in that this method may further comprise the steps:

Step 1: row is selected circuit reference line and following reference line on the gating successively, and last reference line and following reference line are preserved output data behind row reading circuit;

Step 2: set i=1;

Step 3: row is selected the capable pixel of circuit gating i, obtains the data of this row through row reading circuit, and calibration circuit calculates the capable offset of this line data through the output data of row reading circuit in the step 1 simultaneously; Calibration circuit calculates the row offset of this each column data of row through left reference columns and right reference columns;

The computing formula of row offset is:

$R_{i,j}=\frac{i}{2N}{D}_{\mathrm{up},j}+\frac{N-i}{2N}{D}_{\mathrm{down},j}$

Wherein:

R _{I, j}The capable offset of representing the capable j row of i pixel;

N is the line number of detector array;

D _UpBe last reference line;

D _DownBe following reference line;

D _{Up, j}Be D _UpThe j column data;

D _{Down, j}Be D _DownThe j column data;

The computing formula of row offset is:

$C_{i,j}=\frac{j}{2M}{D}_{\mathrm{left},i}+\frac{M-j}{2M}{D}_{\mathrm{right},i}$

Wherein:

C _{I, j}The row offset of representing the capable j row of i pixel;

M is the columns of detector array;

D _LeftBe left reference columns;

D _RightBe right reference columns;

D _{Left, i}Be D _LeftThe i line data;

D _{Right, i}Be D _RightThe i line data;

Step 4: capable each column data of i is carried out the two-dimensional linear compensation through its corresponding capable offset and row offset, and the data after the compensation that this row respectively is listed as through column select circuit are again read successively;

Step 5: make i=i+1, repeating step 3 and step 4 are read until a frame pixel.

Images