Introduction to the composition of LED display screen point-by-point correction technology

1. Concept of point-by-point correction technology

The current status of LED chip production process determines that even if LED chips are produced in the same batch, there are still considerable differences in the luminous intensity and main wavelength between individual LED chips. For LED display applications, this difference will seriously affect the display quality. The indicators such as photometric, chromaticity and electrical parameters must be classified and screened by spectral separation before they can be applied to the same display screen .

However, the use of spectrophotometry and color separation to solve the problem of inconsistent light and color of individual chips cannot achieve perfect image quality due to insufficient precision, the influence of subsequent process flow, and inconsistent light decay during the aging process. In addition, the display screen that has been used for a period of time will also deteriorate in display quality due to inconsistent light decay and other factors, resulting in "flower screen", which is also beyond the reach of spectrophotometry and color separation.

Therefore, the industry tries to solve this problem from the last process of display manufacturing by using different drivers for different LED light points, which is point-by-point correction.

In the late 1990s, the theoretical prototype of point-by-point correction appeared at home and abroad, and the practical exploration of this technology was launched. However, due to the lack of applicable general data collection tools and technical barriers, the research on this technology has long been discontinuous, unsystematic, and self-contained. There is also a lack of communication, and point-by-point correction lacks a recognized definition.

At present, a more reasonable definition is: point-by-point correction, that is, by collecting the brightness (and chromaticity) data of each light point area on the LED screen , the correction coefficient for each light point (or the coefficient matrix for each pixel) is obtained, and it is fed back to the control system of the display screen. The control system applies the correction coefficient to achieve differential drive of each light point, thereby greatly improving the pixel brightness (chromaticity) uniformity of the display screen.

2. Point-by-point correction technology composition

From the above definition, we can see that the point-by-point correction technology can be decomposed into the following four parts:

1. Original data collection;

2. Correction data generation;

3. Drive control;

4. Combination of acquisition system and control system;

The following is an analysis and explanation of these four aspects.

2.1 Raw data collection

Raw data acquisition is the first step of point-by-point calibration, the most basic step, and also the slowest and most difficult step to develop. According to the acquisition parameters, it can be divided into brightness data and color data; according to the acquisition object, it can be divided into module-level acquisition, cabinet-level acquisition, and full-screen area acquisition; according to the acquisition environment, it can be divided into factory mode acquisition and field mode acquisition;

From the perspective of the technical routes and tools collected, it can be roughly divided into the following directions:

1. Mechanical device + photometric probe: that is, the photometric probe is controlled by a mechanical transmission device to collect the number of each light point one by one. In the early experimental device, the screen was placed vertically on the ground, and the luminance meter was moved at equal intervals by a rack to measure point by point . Later, it gradually developed into a machine form, with modules or unit boards placed horizontally and probes collecting data vertically. To improve efficiency, a single machine can be equipped with multiple probes. The author has seen that the maximum number of probes on a single machine is 16, and the data is collected in units of boxes.

The advantage of this acquisition method is its high accuracy, but it also has a fatal flaw: low efficiency. It is difficult to achieve large-scale industrial application. In addition, it is impossible to achieve on-site calibration. In recent years, with the advancement of technology, this machine-based acquisition method is gradually fading out of the historical stage.

2. Digital camera: Using the grayscale data of the light point imaged by the digital camera to achieve point-by-point correction is the cheapest acquisition solution at present. Since 2008, several major display control system manufacturers have invested heavily in research and development, developed their own camera acquisition systems, and carried out point-by-point correction practices, which greatly promoted the promotion and popularization of point-by-point correction technology.

The advantage of digital camera solutions is that the equipment is relatively cheap, but the disadvantages are low accuracy, poor stability, and large differences in consistency between individuals, which makes it difficult to meet the needs of large-scale industrial production. In addition, digital camera solutions are mostly independently developed by control system manufacturers based on their own systems and are incompatible with each other.

3. CCD-based planar brightness/chromaticity distribution measuring instruments: The development of such instruments has been accompanied by the rapid growth of the global flat panel display industry. Using the principle of imaging brightness measurement, it can efficiently obtain the brightness/chromaticity value of any area on the imaging plane. Since 2006, related products have been launched in Japan, the United States, Denmark, France, Germany and China, but few can meet the special requirements of LED point-by-point correction.

This type of equipment has high precision, good stability and excellent correction effect, but is relatively expensive.

4. Industrial CCD acquisition solution: In addition to the above directions, there are also some solutions based on industrial cameras, such as the industrial camera module calibration pipeline solution independently developed by Barco; another example is Changchun Xida, a pioneer in point-by-point calibration. Their independently developed and continuously improved industrial CCD calibration solution is the first integrated bright and color calibration solution in China.

If you want to do your work well, you must first sharpen your tools. With the improvement of the efficiency and enhancement of the functions of the acquisition tools, point-by-point correction data acquisition has a wider space and possibility, extending from the factory to the site, from the new screen to the old screen, from the flat screen to the curved screen and even the special-shaped screen.

2.2 Generation of calibration data

The generation of correction data can be divided into three parts: the first is the correction processing of raw data, the second is the setting of the correction target value, and the third is the calculation and generation of correction data. The most important technical breakthrough lies in the "correction processing of raw data", especially the data correction in the on-site correction environment.

2.2.1 Correction of original data

The simplest case of on-site correction is: for a flat screen, the best audience area of ​​the display screen is selected as a single data collection position, and data is collected in each area of ​​the full screen in turn. The data collected in this way will inevitably contain systematic errors introduced by different observation angles. The collected data shows that the brightness is high in the vertical normal direction, and the brightness decreases in the direction deviating from the normal direction. The greater the deviation angle, the lower the brightness. If it is not corrected, the calibrated display screen will inevitably be dark at the bottom and bright at the top; dark in the vertical direction of the position and bright on both sides; when viewed away from the correction point, the brightness and darkness will be distorted.

However, when the screen is curved or the on-site environment is limited and multiple cameras are required to complete the acquisition, due to the different acquisition angles of different cameras, if no correction is made, there will be obvious dividing lines at the seams.

The above problems make it impossible to calibrate many screens on site. Recently, some digital camera solutions use the neighboring area comparison feedback method, and some devices use the full screen image as a reference method for correction.

2.2.2 Setting of calibration target value

The setting of the calibration target value is also a part of the point-by-point calibration technology that deserves in-depth discussion. As we all know, brightness calibration loses brightness, and chromaticity calibration loses both brightness and color gamut space and color saturation. So how to set reasonable calibration target brightness and chromaticity values, and find the best balance between brightness, color gamut and uniformity in combination with customer needs?

At present, many digital camera calibration solutions put the target value setting link before acquisition due to the lack of intermediate data. However, different displays have different optimal balance points, especially for chromaticity calibration. Improper target value setting will directly lead to calibration failure! Reasonable target value setting depends on statistical analysis of collected data. Therefore, we put the target value setting after the acquisition is completed, and provide various auxiliary parameters and graphs to help users adjust the target value.

2.3 Drive Control

With the correction data, the correct application of the control system is required to achieve point-by-point correction.

There are two ways to achieve drive control: one is current amplitude control, and the other is pulse width control ( PWM mode). Since the current amplitude and brightness are not strictly linear, and the increase or decrease of current will cause the main wavelength of the LED chip to shift, current control is used less and less. The main way to achieve point-by-point correction drive control is to adjust the pulse width.

Major domestic control system suppliers have already realized point-by-point differential drive control of LED lights. However, due to the lack of universal acquisition equipment, point-by-point correction was still a unique technical advantage of a few industry leaders with their own control systems until 2008. With the breakthrough progress of acquisition equipment, the point-by-point correction function of the control system, which was mostly a promotional selling point in 2008 and could not be put into practical use, has gradually become a necessary tool for control systems to enter the market by 2010. Today, there are very few full-color display control systems on the market that do not have the ability to correct brightness point by point.

However, there are still areas to be improved in the drive control of point-by-point correction, as shown in the following aspects:

1. The low brightness and linearity of the calibration need to be improved;

2. Currently, there are only a few systems that have color correction capabilities;

3. The number of load points after calibration needs to be expanded;

In addition to using the control system to achieve drive control, there is another technical idea to achieve correction from the signal source level by real-time processing of the front-end video stream. It can be divided into hardware implementation and software implementation. Hardware implementation is to add a signal processor between the video signal source and the control system, store the correction data internally, apply the correction data to the input video stream signal for real-time calculation, and then output it to the control system. Software implementation is to intercept the display data stream with the computer as the signal source, perform correction data calculation, and then output it to the DVI port.

Compared with the correction achieved by the control system, since the DVI signal is only 8 bits, this method of using the front-end video processor to achieve correction will cause serious loss of grayscale, and its low brightness and poor linear performance will be the inevitable result. When applying chromaticity correction, the effect will also be unsatisfactory due to insufficient precision.

2.4 Integration of data acquisition system and control system

The point-by-point calibration process requires the following three steps: the control system controls the screen to display red, green and blue images in the specified area; the acquisition system completes the acquisition respectively; and the calibration data is generated and written into the control system.

Before 2009, most acquisition systems were developed by the control system itself and used in conjunction with its own control system. Therefore, whether it is module-level, cabinet-level or full-screen-level correction, the acquisition system and the control system mostly use a custom control interface protocol to interact.

This combination of control interface protocols has a high degree of automation and the process of writing data does not require manual intervention. For some digital camera calibration solutions, due to the lack of accuracy and stability of the measurement equipment, repeated acquisition and neighboring area comparison are required to ensure consistency between regions. Therefore, close interaction between acquisition and display control is required. This combination of control interface protocols is the only option. However, this also causes the acquisition system and the control system to be bundled and incompatible with each other. At a time when control system technology is constantly innovating and developing, and new control systems are constantly emerging, it is undoubtedly not in the interests of LED screen manufacturers.

There are two situations when LED screen manufacturers introduce imported acquisition equipment and combine it with their own control system. One is to modify the control system in accordance with the control interface protocol requirements of the acquisition system and use the software functions of the acquisition system to complete the correction; the other is to develop software on their own to achieve display control, acquisition system acquisition, and the generation and writing of correction data. Both of these situations mean that the technology is difficult to introduce and the cost is high. Similarly, it is incompatible and cannot be combined with the common control system in the market.

Because the acquisition equipment is professional and stable, the calibration can be completed with only one data acquisition, so the display control part becomes very simple and can be completed without interactive communication with the control system . As for the step of writing to the control system, the data file can be used, and the control system can complete the work of reading and writing related storage areas by itself.

In this way, the general control system does not need to be modified or expose any control interface commands. It can achieve point-by-point correction by reading the correction data document in the open format of Zhongkeweiyou. The workload of system docking is compressed to a minimum, and the acquisition system also achieves maximum compatibility with emerging control systems at the lowest cost.

The combination of data interface protocols is simple, flexible and compatible, but has a low degree of automation. In the future, the formation of a standard point-by-point correction control interface protocol in the industry will be the ultimate solution to this problem.