CCM imaging quality index of camera module
August 23, 2021
CCM imaging quality index of camera module
1: Imaging uniformity (Corner Shading)
Test environment: whiteboard test environment
Due to the relationship of the camera lens, the light energy obtained by each part of the image sensor surface is different, which leads to the difference in brightness and color of the four corners or the periphery and the center of the image, that is, the unevenness of the image, as shown in the figure.
1.1Brightness uniformity
1.2 Color difference
2: Geometric distortion
Test environment: Distortion test card test environment.
Geometric distortion, or distortion, refers to a kind of aberration that loses the similarity of the object image caused by the change of magnification with the increase of the field of view, which only affects the shape of the image and does not affect the sharpness of the image.
The position on the image plane is different, and there are different magnifications. Therefore, object deformation occurs, which is divided into barrel distortion and pincushion distortion.
2.1 Algorithm One
2.2 Algorithm 2
This method is actually based on the principle of image correction. The distorted image is corrected according to the mathematical mapping relationship between the distorted image and the corrected image:
The mapping from any point (x, y) in the distorted image to the position (X, Y) of the corrected image
Distortion center position (x0, y0)/(X0, Y0), at this time x0=X0, y0=Y0;
Pixel ratio of the corrected image K (mm/pixel)
3: Image color restoration
Test environment: color test card test environment.
3.1 RGB color model
3.2 HSV color model
Color can be described by hue/H (hue/Hue), saturation/S (purity/Saturation) and brightness/V (lightness/Value).
Any colored light seen by the human eye is the combined effect of these three characteristics, which are the three elements of color. The hue is directly related to the wavelength of the light wave, and the brightness and saturation are related to the amplitude of the light wave.
The color saturation represents the vividness of the color of the playback light, which depends on the grayscale in the color. The higher the grayscale, the lower the color saturation, and vice versa.
RGB and HSV color space model
The hue H parameter represents color information, that is, the position of the spectrum color. This parameter is represented by an angle, and red, green, and blue are separated by 120 degrees. The complementary colors are 180 degrees apart.
The purity S is a ratio value ranging from 0 to 1, which is expressed as the ratio between the purity of the selected color and the maximum purity of the color. When S=0, only gray scale.
The lightness V represents the brightness of the color, ranging from 0 to 1. One thing to note: There is no direct relationship between it and light intensity.
3.3 Lab color space
The RGB mode is an additive color mode for light-emitting screens, while the Lab mode does not rely on light or pigments. It is determined by the International Commission on Illumination (CIE) to theoretically include all colors that can be seen by the human eye. Color mode.
It uses L*, a*, and b* three mutually perpendicular coordinate axes to represent a color space. The L* axis represents lightness, with black at the bottom and white at the top; +a* represents magenta, -a* represents green; +b* represents yellow, and -b* represents blue. The a* axis is the red-green axis, and the b* axis is the yellow-blue axis.
Lab has three channels, making a Lab mode image, from which you can find that the a and b channels are almost indistinguishable. They only define the color part of the image, but not the morphological details of the image. Therefore, the color in the Lab mode image It is separate from the level, which is different from RGB. This allows the gray curve and the color curve to be adjusted separately. When the user adjusts the gray curve, the color part is not affected, so you can only use the L without color information. *The channel determines the brightest and darkest values of the image.
In this way, you can avoid some jumpy and discontinuous colors during the color correction process, and you can also sharpen the image only on the L channel to emphasize the overall subtle level of the image.
The L component in the Lab color space is used to represent the brightness of the pixel, and the value range is [0,100], which means from pure black to pure white; a means the range from red to green, and the value range is [127,-128]; b represents the range from yellow to blue, and the value range is [127,-128].
3.4 Color reproduction analysis
In order to test the color restoration ability of the camera imaging, the 24-color standard color card was taken under different color temperature/light source conditions, and the average R, G, B of each color block was analyzed, and it was converted to the Lab color space. The standard color of the card is compared and calculated:
Name restoration error, chromaticity restoration error, total chromaticity restoration error and total color restoration error.
4: White Balance
Test environment: whiteboard test environment.
Color temperature represents the components of the spectrum, the color of light. A low color temperature indicates that there are many long-wave light components; when the color temperature changes, the ratio of the three primary colors in the light source will change, and the ratio of the three primary colors needs to be adjusted to achieve the color balance. This is the white balance adjustment technology.
In the environment of different color temperature, the image should be white if you photograph a white object
4.1 White balance value display
Intercept an area in the image area, calculate the RGB average value R, G, B of the interval, and its white balance is:
Under different color temperature environments, when lighting white or neutral gray objects, when the ratio of the three primary colors reaches the color balance, the White Balance value is close to 0; when it exceeds a certain value greater than 0, it indicates that the white balance is poor.
4.2 White balance RGB graphic representation
Also in different color temperature environments, when illuminating a white or neutral gray object, follow a specified path (such as a straight line) in its imaging, the path point is the X coordinate, and the R, G, and B values of each pixel on the path are The three graph lines formed by the Y coordinate indicate that the three curves overlap together, indicating that the RGB colors are balanced.
5: Dynamic range
Test environment: gray-scale card test environment.
Dynamic range expresses the ability of an image sensor to capture both light and dark objects in the same photo. It is usually defined as the logarithm of the ratio of the brightest signal to the darkest signal (noise threshold level), and 54dB is usually used as the general purpose of commercial image sensors. index.
An image sensor with a wider dynamic range can provide better performance in bright light environments (for example, photos taken with a narrower dynamic range sensor in bright light will appear "washed" or blurred.
5.1 Grayscale description
The background color of the gray scale test chart is medium gray. Under the approximate condition of taking an integer, 20 gray scales out of a total of 256 gray scales from RGB (0, 0, 0) to RGB (255, 255, 255) are uniformly extracted. Use 20 rectangular blocks with the same area to fill the above-mentioned 20 levels of gray respectively.The reflection density of each level differs by 0.1. The reflection density corresponding to the three calibrations of A, M and B in the chart is 0.05, 0.75 and 1.65. They represent highlights, neutral grays and shadows. The background density is the same as the M point.
5.2 Grayscale curve description
6: Imaging resolution (Resolution)
Test environment: definition card test environment.
First, in general, the shooting distance when testing the resolution is equal to the focus distance of the module, that is, when testing the resolution, the module is placed at the focus distance to shoot the test target.
The test card usually adopts the ISO12233 test card or other customized test cards, and the measurement algorithm generally adopts MTF, TV_Line, SFR.
6.1 TV_Line(Transmission Line Pulsing)
The full name of TV Line is Transmission Line Pulsing, also known as "fountain map", unit: "line number". It means that from the horizontal direction, it is equivalent to erecting each row of scanning lines, and then multiplying the aspect ratio to form a horizontal bus, which is called the horizontal resolution.The difference between image composition and pixels: TV Line means TV line, which is used to measure the ability to "decompose the details of the subject"; pixels are small dots of various colors. The more pixels, the more the image can be placed. The bigger the clearer. Both can form an image, but the principle of forming an image is different: TV Line is imaged by lines, while pixels are imaged by dots.
TV line is mainly used for subjective testing. Due to the wide application of computers, the number of TV lines is also detected by image processing software.
Image processing algorithm
Scan line by line in the direction of TV Line, analyze the gray level curve of each line, find the first derivative and second derivative respectively, and determine the boundary of the line according to the change of the derivative, that is, "count out" the number of lines at this time; Scan, analyze and process, and "number lines" until the "number lines" cannot be counted.
Using the "fountain map" to measure is more intuitive, but it is related to whether the test card is full in the image area, and the environmental requirements are slightly higher. There are more Japanese and Korean manufacturers and end customers.
6.2 MTF (Modulation Transfer Function)
MTF is the abbreviation of Modulation Transfer Function in English, and the Chinese is Modulation Transfer Function. The function that the modulation degree changes with the spatial frequency is called the modulation degree transfer function, which describes the ability to distinguish spatial details, and the unit is expressed in line/mm.
The modulation transfer function is similar to the concept of contrast, and its mathematical expression is as follows:
It represents the change in contrast after shooting with the camera. If MTF=1, it means that the image quality is very high and the transmission is basically lossless. It measures the transfer ability of the lens to transform the original image into the specific image.
In the actual test process, the wire pairs shown below are usually used for analysis and calculation.
In the above figure, the top is the original image, and the bottom is the image taken by the camera. It can be seen that the edge of the line pair is obviously blurred after passing through the lens, and the denser and more blurred the line pair, the lower the corresponding MTF.
The MTF test is related to whether the test card is full in the image area, which is used by domestic CCM manufacturers;
6.3 SFR (Spatial Frequency Response)
FR is the abbreviation of Spatial Rrequency Response, which is mainly used to measure the impact on a single image as the line of spatial frequency increases. In short, SFR is another way of expressing MTF.
The test methods include slit, pinhole test and bevel test. However, the hypotenuse test is usually used. This method simplifies the test process to a large extent and does not require high test environment.
When expressing camera image resolution, MTF50 or MTF50P is usually used.
MTF50 represents the corresponding spatial frequency when MTF is 50% of the maximum value (ie, MTF=0.5).
In the process of image shooting, the image processing module may sharpen the image, which will affect the value of MTF. In order to avoid such an impact, it is stipulated that MTF50P is 50% of the maximum MTF of the image taken by the camera. Corresponding spatial frequency.
SFR is relatively abstract, and it is used by European and American manufacturers and end customers.
6.4 Summary
Therefore, when selecting the MTF and TV_Line algorithms, it is necessary to accurately calculate the size of the test card according to the field of view of the camera module, so that the image area of the test card is just full of the screen. The reading at this time needs to be converted into the correct resolution value.
Due to the lens, different positions of the resulting image have different image definitions. Therefore, when measuring, choose different image positions. For TV_Line and SFR, it is even to be measured separately along the vertical and horizontal directions.
MTF and TV_Line are related to whether the test card is full in the image area, and the environmental requirements are slightly higher;
SFR has nothing to do with whether the test card is full in the image area and does not require high environmental requirements.
7: Optical image stabilization test
Test environment: definition card test environment.
When simulating the jitter mechanism of the camera module, the position measurement of the target point on the test card is repeated multiple times, and the position change of the target point is calculated to evaluate the effect of optical anti-shake.
7.1 Test model
7.2 Test process