CN114584757A

CN114584757A - Simple CCD full-well testing method

Info

Publication number: CN114584757A
Application number: CN202111500957.9A
Authority: CN
Inventors: 常振; 王煜; 林方; 赵欣; 司福祺
Original assignee: Hefei Institutes of Physical Science of CAS
Current assignee: Hefei Institutes of Physical Science of CAS
Priority date: 2021-12-09
Filing date: 2021-12-09
Publication date: 2022-06-03

Abstract

The invention discloses a simple CCD full-well testing method, which comprises the following steps: driving the LED to emit light by a voltage-adjustable power supply to obtain point light sources with different brightness; acquiring an LED image through a common civil lens; controlling the position of the LED in the CCD pixel through a two-dimensional displacement platform; the method also comprises a data acquisition method: adjusting the position of the LED relative to the CCD to enable the LED light spot irradiation range to be about 9 pixels; obtaining a pixel overflowing firstly by analyzing the DN value of each pixel along with the increase of the LED brightness; by analyzing the DN value change of the central pixel and the DN value of the first overflow pixel, the DN value increase inflection point of the first overflow pixel is obtained by a nonlinear interpolation and curve fitting method, the DN value of the central pixel corresponding to the point is the full-trap DN value, and the full-trap numerical value of the CCD pixel can be calculated by combining with the system gain.

Description

Simple CCD full-well testing method

Technical Field

The invention relates to the technical field of CCD imaging, in particular to a simple CCD full-well testing method.

Background

During the development of the CCD imaging circuit, it is usually necessary to test the full-well value many times to verify the working performance of the circuit. The PTC method is used for measurement every time, the workload is huge, and the development efficiency is low. Thus: 1. in order to improve the research and development efficiency of the CCD imaging circuit, the LED point test method can greatly simplify the test process and improve the development efficiency; 2. aiming at the full-well test of a single pixel, a commonly used absolute point light source system is complex and has high cost, and in order to simplify the test cost, an LED point test method can be used; 3. theoretical analysis and experimental verification prove that the method has high reliability, and the research and development efficiency can be greatly improved by using the method as a full-well testing method in the research and development process of the CCD imaging circuit.

For testing the full-well performance of a CCD imaging system, a Photon Transfer Curve (PTC) method is generally used, and the PTC method generally requires building a flat-field surface light source (for example, using an integrating sphere and other devices), and the structure of the testing system is complex. In fact, because the CCD is very sensitive to the response of light, saturation occurs when the CCD is illuminated with the lowest luminance of a general integrating sphere, and a customized low-illumination integrating sphere is required for testing. The PTC method is premised on the fact that the light source is an absolute flat field, because the method actually uses space dimension noise as time dimension noise, all pixels are required to receive light with the same brightness, and the requirement on stray light of a test environment is higher; an ideal curve can not be obtained sometimes by using the PTC method, particularly for scientific-grade CCDs, due to the fact that the PTC is abnormally bent in a linear region due to factors such as the influence of pixel correlation, and the deviation is increased when the bent PTC curve is used for calculating the full-well parameter; because the PTC method adopts a statistical method to calculate the CCD full well, the full well value of a single pixel cannot be obtained.

Although the CCD is marked with the full-well parameters when it leaves the factory, since the design parameters of the circuit, such as the frequency of the horizontal or vertical transfer clock, the driving current, etc., may affect the full-well performance of the CCD, it is necessary to continuously adjust the parameters and test the full-well of the CCD during the development of the imaging circuit to find the parameter combination corresponding to the optimal full-well performance. Dozens of times or even hundreds of full-well tests are sometimes required in the parameter adjustment process, and the PTC method used in each test is low in efficiency.

Specifically, for testing the full-well performance of a CCD imaging system, a Photon Transfer Curve (PTC) method is generally used, and important parameters such as CCD readout noise, system gain, full-well charge number, pixel response non-uniformity (PRNU), and the like can be obtained by building a flat-field light source. The method uses uniform light with different intensities to irradiate a CCD focal plane, performs data processing on the acquired image, and calculates to obtain a signal-variance curve of the CCD response value under different light intensities, wherein the curve is a photon transfer curve. In addition, CCD response values of different levels can be obtained by changing the exposure time of the CCD device. The photon transfer curve can be used to test whether the CCD is in an optimal operating state. The horizontal axis of the curve is the mean value of the CCD signal response values, and the vertical axis is the noise of the CCD in the state, and can be expressed by the standard deviation or the variance of the CCD response values. The PTC technology is a relatively well-known method for testing multiple parameters of the CCD and originates from foreign countries.

The EMVA1288 standard is a set of standards for a set of unified camera and graphic sensor performance tests officially released by the European Machine Vision Association (EMVA) in 8 months of 2005. According to the definition of the EMVA1288 standard, the saturation radiation amount is a radiation amount value corresponding to a gray value variance (DN value) and the maximum value of a radiation amount relation curve, and the saturation capacity can be calculated by combining the conversion relation between the radiation amount and the electron Number, namely quantum efficiency. However, the standard does not define a Full Well Capacity (FWC) calculation method, which is limited by the number of analog-to-digital conversion bits, and the full well capacity is generally higher than the saturation capacity. When the full-well output of the CCD does not reach the range of analog-to-digital conversion (ADC) of the imaging system, the full-well charge number can be calculated by adopting the EMVA1288 standard definition. Full wells can be calculated from either a 3% or 5% linear fit deviation as the saturation point, but this human selection factor is likely to give large deviations to the test results.

In summary, the prior art has the following technical problems:

the PTC method is premised on the fact that the light source is an absolute flat field, because the method actually uses space dimension noise as time dimension noise, all pixels are required to receive light with the same brightness, and the requirement on stray light of a test environment is higher; an ideal curve can not be obtained sometimes by using the PTC method, particularly for scientific-grade CCDs, due to the fact that the PTC is abnormally bent in a linear region due to factors such as the influence of pixel correlation, and the deviation is increased when the bent PTC curve is used for calculating the full-well parameter; because the PTC method adopts a statistical method to calculate the CCD full well, the full well value of a single pixel cannot be obtained.

The conventional full well test rule for calculating the full well charge number by using EMVA1288 standard definition is to determine the DN value of the full well based on artificial definition, and the artificial selection factor is likely to bring large deviation to the test result.

When measuring a single-point full trap, an absolute point light source system can be constructed, but equipment such as a collimator, a star point plate and a lens matched with a numerical aperture is needed, so that the cost is high.

Disclosure of Invention

The simple CCD full-well testing method provided by the invention can solve the technical problems.

In order to achieve the purpose, the invention adopts the following technical scheme:

a CCD full-well test method comprising:

firstly, a test system is set up, wherein the test system comprises an imaging system, a CCD (charge coupled device) and an LED point light source, the imaging system is connected with a computer, and the LED point light source covers a 3 x 3 pixel area in a CCD image by adjusting a two-dimensional displacement platform in combination with a common imaging lens;

the test method comprises the following steps,

1) adjusting the voltage of the LED to V₁The LED is lightened, and the focal length of the lens and the distance and the position between the LED and the detector are adjusted, so that the LED is focused into a bright spot on a CCD image and occupies a pixel area of 3 multiplied by 3;

2) observing DN value distribution corresponding to the LED, adjusting the position of the DN value distribution to ensure that the DN value distribution accords with the following rule, wherein the rule is that the DN values of pixel points in the 9 pixel point position distribution are defined as upper left, middle left and right, middle left, right, lower left, middle left and right, and upper left, middle left, right, middle upper and lower middle pixels are close to each other, N images are obtained, N is a natural number larger than 1, and the average value of the DN values of the pixel points in the N images is calculated;

3) sequentially changing the voltage of the LED into Vi, obtaining different light intensity levels, repeating the step 2), obtaining data of 9 pixel points under different light intensities, and respectively recording the data as ViP_1i、P_2i、……、P_9iI is 1 to M, and M is the number of all light intensity levels;

4) taking voltage Vi as a horizontal axis and pixel DN value P_1i、P_2i、……、P_9iNine curves are drawn for the longitudinal axis, the area range near the curve inflection point is selected, a polynomial fitting method is adopted, a key curve, namely the curve inflection point of the overflow pixel DN value and the corresponding central pixel DN value are obtained through calculation, and the value is the DN value corresponding to the central pixel when the central pixel reaches the full well.

According to the technical scheme, the CCD full-well testing method comprises the following steps: driving the LED to emit light by a voltage-adjustable power supply to obtain point light sources with different brightness; acquiring an LED image through a common civil lens; controlling the position of the LED in the CCD pixel through a two-dimensional displacement platform; the method also comprises a data acquisition method: adjusting the position of the LED relative to the CCD to enable the LED light spot irradiation range to be about 9 pixels; obtaining a pixel overflowing firstly by analyzing the DN value of each pixel along with the increase of the LED brightness; by analyzing the DN value change of the central pixel and the first overflowing pixel, the DN value increasing inflection point of the first overflowing pixel is obtained by a nonlinear interpolation and curve fitting method, the DN value of the central pixel corresponding to the point is the full-trap DN value, and the pixel full-trap numerical value can be obtained by combining the system gain.

According to the invention, a simple LED point light source system is constructed, the DN value of the pixel full well of the CCD can be obtained by combining numerical analysis, and the charge number of the pixel full well can be obtained by combining system gain.

In the research and development process of the CCD imaging circuit, in order to ensure that the CCD works at the best full-well performance, the value and the change rule of the full-well of the CCD need to be tested continuously. The conventional full trap test method has complex system and inconvenient test. The invention can simplify the testing process and improve the research and development efficiency of the CCD imaging circuit.

Drawings

FIG. 1 is a block diagram of an LED spot test method system of the present invention;

FIG. 2a is a pixel distribution of LED dot DN values for a schematic of the LED dot test method of the present invention;

FIG. 2b is a DN value versus light intensity curve of a schematic diagram of an LED spot test method of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention.

In order to simplify the testing process, the embodiment of the invention uses an LED point testing method, namely an LED is used as a point light source with adjustable light intensity, a common imaging lens is combined to enable the LED point light source to cover a 3 x 3 pixel area in a CCD image, and a full-well numerical value can be obtained by combining a numerical value calculation method according to the pixel value change rule of the LED image.

The method comprises the following specific steps:

the CCD full-well testing method, namely the LED point testing method, disclosed by the embodiment of the invention is set up as shown in figure 1. Namely, an LED is used as a point light source with adjustable light intensity, and a common imaging lens is combined, so that the LED point light source covers 3 multiplied by 3 pixel areas in a CCD image by adjusting a two-dimensional displacement platform.

Normally, the CCD will overflow in the horizontal or vertical adjacent image element first, and the DN values of the horizontal or vertical three image elements in the LED point light source image are as shown in fig. 2. Fig. 2a is a schematic diagram of imaging of an LED point in three pixels horizontally or vertically, and fig. 2b is a schematic diagram of a change rule of DN values of the three pixels as light intensity increases.

The region <1> in fig. 2b is a linear growth region, at this time, all three pixels do not reach the full well, and as the light intensity increases, the DN values of the three pixels are synchronously enhanced, and the specific rule should conform to the growth rule of a gaussian curve; when the pixel b in the figure is close to a full well, the capacity of collecting charges is weakened, the possibility that photo-generated charges are captured by adjacent pixels is increased, the increasing curve of the pixel b is gradually slowed down, the increasing curve of the adjacent pixels starts to be accelerated, and the pixel (a) on one side of the reading direction usually enters the region <2> earlier, namely the full well region; the light intensity is increased continuously, the b pixel reaches the saturation value and does not rise any more, and then the a and c pixels also reach saturation, namely the area <3> in the figure.

In the full-trap area, polynomial fitting is carried out on the DN value variation trend of the pixel which is influenced by overflow firstly in the image element shown in the image element 2(b), the inflection point of the pixel curve shown in the image element 2(b) can be obtained according to the result of the polynomial fitting curve, and the full-trap charge number of the CCD at the pixel element at the point can be obtained by combining the system gain according to the DN value corresponding to the inflection point of the full-trap area.

Specifically, the method is simple and rapid in data acquisition and can be used as a supplement of the PTC method. The specific steps for acquiring data are as follows:

2) observing DN value distribution corresponding to the LED, adjusting the position of the DN value distribution to ensure that the DN value distribution conforms to the rule in the figure 2(a), namely, 9 pixel point azimuth distribution is defined as upper left, middle and right, middle left, middle, right, lower left, middle and right, and middle left, right, middle left, middle upper and lower pixel DN values are similar, namely, the left pixel element DN value and the right pixel element DN value are similar, and the upper pixel element DN value and the lower pixel element DN value are similar, acquiring 16 images, and calculating the average value of the DN values of the 9 pixel points in the 16 images;

3) sequentially changing the voltage of the LED to Vi, obtaining different light intensity levels, repeating the step 2), obtaining data of 9 pixel points under different light intensities, and respectively recording the data as P_1i、P_2i、……、P_9i(i 1-M, M being the number of all intensity levels);

4) taking voltage Vi as a horizontal axis and pixel DN value P_1i、P_2i、……、P_9iDrawing three curves for the longitudinal axis, selecting a proper range, and calculating by adopting a polynomial fitting method to obtain an inflection point of a key curve (an overflow pixel DN value curve at first) and a corresponding central pixel DN value, wherein the inflection point is the DN value corresponding to the central pixel when the central pixel reaches the full well.

In summary, the CCD full-well testing method of the embodiment of the present invention includes system construction, that is, driving the LEDs to emit light by the voltage-adjustable power supply to obtain point light sources with different brightness; acquiring an LED image through a common civil lens; controlling the position of the LED in the CCD pixel through a two-dimensional displacement platform; the method also comprises a data acquisition method: adjusting the position of the LED relative to the CCD to enable the LED light spot irradiation range to be about 9 pixels; obtaining a pixel overflowing firstly by analyzing the DN value of each pixel along with the increase of the LED brightness; and (3) obtaining a DN value increase inflection point of the first overflowing pixel by analyzing the DN value change of the central pixel and the first overflowing pixel through a nonlinear interpolation and curve fitting method, wherein the DN value of the central pixel corresponding to the point is the full-trap DN value.

The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A simple CCD full-well testing method is characterized in that,

the test method comprises the following steps,

1) and adjusting the voltage of the LED to V₁The LED is lightened, and the focal length of the lens and the distance and the position between the LED and the detector are adjusted, so that the LED is focused into a bright spot on a CCD image and occupies a pixel area of 3 multiplied by 3;

2) observing DN value distribution corresponding to the LED, adjusting the position of the DN value distribution to ensure that the DN value distribution accords with the following rule, wherein the rule is 9 pixel points, the azimuth distribution is defined as upper left, middle and right, middle left, middle and right, lower left, middle and right, middle left, right, upper middle and lower middle pixel DN values are similar, N images are obtained, N is a natural number larger than 1, and the average value of the DN values of the 9 pixel points in the N images is calculated;

3) sequentially changing the voltage of the LED to Vi to obtain different lightsAnd (4) repeating the step 2) to obtain data of 9 pixel points under different light intensities, and respectively recording the data as P_1i、P_2i、……、P_9iI is 1 to M, and M is the number of all light intensity levels;