CCD and CMOS sensors have been in constant competition over the past few years. In this article, we will try to consider the advantages and disadvantages of these technologies. CCD-matrix (abbreviated from "charge-coupled device") or CCD-matrix (abbreviated from CCD, "Charge-Coupled Device") is a specialized analog integrated circuit consisting of light-sensitive photodiodes, made on the basis of silicon, using CCD technology - charge coupled devices. In a CCD sensor, the light (charge) incident on the sensor pixel is transmitted from the microcircuit through one output node, or through just a few output nodes. The charges are converted to a voltage level, accumulated and sent out as an analog signal. This signal is then summed and converted to numbers by an analog-to-digital converter outside the sensor. CMOS (complementary logic on metal-oxide-semiconductor transistors; CMOS; English CMOS, Complementary-symmetry / metal-oxide semiconductor) is a technology for constructing electronic circuits. In the early stages, conventional CMOS chips were used for display, but the picture quality was poor due to the low light sensitivity of the CMOS elements. Modern CMOS sensors are manufactured using more specialized technology, which has led to a rapid increase in image quality and light sensitivity in recent years. CMOS chips have several advantages. Unlike CCD sensors, CMOS sensors contain amplifiers and analog-to-digital converters, which significantly reduces the cost of the final product, because it already contains all the necessary elements to obtain the image. Each CMOS pixel contains electronic converters. CMOS sensors have more functionality and wider integration capabilities. One of the main problems when using CMOS sensors in video cameras was the image quality. CCDs have provided and are now providing lower noise levels. As a result, CMOS chips behave extremely poorly in low light compared to CCD chips. And since low light is one of the main difficulties in video shooting, this was the main barrier to using CMOS sensors. However, the manufacturing experience gained over the years of CMOS development has made it possible with each new generation of these sensors to significantly reduce the fixed and random noise that affects the image quality. Another weak point of CMOS is the distortion that appears when capturing a dynamic image due to poor sensor sensitivity. Car images may contain very bright elements such as headlights, the sun, or very dark areas such as on license plates. For this reason, a wide dynamic range is required to handle scenes with large contrast drops. The CCD sensor has good dynamic range parameters, but the CMOS access to individual pixels gives much more opportunities to obtain a better dynamic range. Also, when using CCDs, bright spots in the scene can create vertical lines in the picture and interfere with license plate recognition due to fading and blurring. Despite the fact that CCDs have a higher sensitivity characteristic, the main factor limiting their use is the low rate of charge readout and, as a result, the inability to provide a high speed of imaging. The higher the matrix resolution, the slower the imaging speed. In turn, CMOS technology, combining a photosensitive element and a processing microcircuit, allows obtaining a high frame rate even for 3 megapixel sensors. However, the use of megapixel CMOS sensors for IP cameras in video surveillance systems requires efficient compression of the data stream. The most common IP CCTV compression algorithms currently are M-JPEG, MPEG4 and H.264. The first is often implemented directly on the CMOS sensor by the matrix manufacturer itself. MPEG4 and H.264 algorithms are more efficient, but require a powerful processor. To form a real-time stream with a resolution of more than 2 megapixels, CMOS IP cameras use coprocessors that provide additional calculations. Currently, IP cameras based on CMOS sensors are becoming more and more popular primarily due to the technology support from the leaders of IP video surveillance. At the same time, their cost is higher than similar cameras on CCD. This is despite the fact that CMOS technology, combining the analog and digital parts of the device, allows you to create cheaper cameras. The situation is such that today the cost of an IP camera is determined by its capabilities and characteristics. The main thing is not the type of matrix, but the software implemented by the camera processor.

Advantages of CCD matrices: Low noise level, high pixel fill factor (about 100%), high efficiency (the ratio of the number of registered photons to their total number hitting the light-sensitive area of \u200b\u200bthe matrix, for CCDs - 95%), high dynamic range (sensitivity), good sensitivity in the IR range.

Disadvantages of CCD matrices: The complex principle of reading the signal, and therefore the technology, the high level of power consumption (up to 2-5W), is more expensive to manufacture.

Advantages of CMOS matrices: High speed (up to 500 frames / s), low power consumption (almost 100 times compared to CCD), cheaper and easier to manufacture, promising technology (on the same crystal, in principle, it costs nothing to implement all the necessary additional circuits : analog-to-digital converters, processor, memory, thus obtaining a complete digital camera on one chip).

Disadvantages of CMOS matrices: Low pixel fill factor, which reduces sensitivity (effective pixel surface is ~ 75%, the rest is occupied by transistors), high noise level (it is caused by the so-called tempo currents - even in the absence of illumination, a fairly significant current flows through the photodiode), the fight against which complicates and increases the cost of technology, low dynamic range.

Like any technology, CMOS and CCD technologies have advantages and disadvantages, which we have tried to consider in this article. When choosing cameras, it is necessary to take into account all the pros and cons of these technologies, paying attention to such parameters as light sensitivity, wide dynamic range, power consumption, noise level, and cost of the camera.

The image sensor is an essential element of any camcorder. Almost all cameras today use CCD or CMOS image sensors. Both types of sensors perform the task of converting the image built on the sensor by the lens into an electrical signal. However, the question of which sensor is the best is still open.

N.I. Chura
Technical advisor
LLC "Microvideo Group"

The CCD is an analog sensor, despite the discreteness of the photosensitive structure. When light hits the matrix, each pixel accumulates a charge or a packet of electrons, which is converted, when read on the load, into a video signal voltage proportional to the illumination of the pixels. The minimum number of intermediate transitions of this charge and the absence of active devices ensure a high identity of the sensitive elements of the CCD.

The CMOS sensor is a digital device with Active Pixel Sensor. Each pixel has its own amplifier, which converts the charge of the sensitive element into voltage. This makes it possible to practically individually control each pixel.

Evolution of CCD

Since the invention of the CCD by Bell Laboratories (or Bell Labs) in 1969, the size of the image sensor has been continuously decreasing. At the same time, the number of sensitive elements increased. This naturally led to a decrease in the size of a single sensitive element (pixel), and, accordingly, its sensitivity. For example, since 1987, these sizes have decreased 100 times. But thanks to new technologies, the sensitivity of one element (and therefore of the entire matrix) has even increased.

What allowed to dominate
From the outset, CCDs have become the dominant sensors because they provide better image quality, less noise, higher sensitivity, and greater pixel uniformity. The main efforts to improve the technology have focused on improving the performance of the CCD.

How sensitivity grows
Compared to the popular Sony HAD standard definition (500x582) sensor of the late 1990s. (ICX055) The sensitivity of the more advanced Super HAD technology has increased almost 3 times (ICX405) and Ex-view HAD - 4 times (ICX255). Moreover, for black and white and color versions.

For high-resolution matrices (752x582), the success is somewhat less impressive, but if we compare the Super HAD color image models with the most modern Ex-view HAD II and Super HAD II technologies, the increase in sensitivity will be 2.5 and 2.4 times, respectively. And this is despite the decrease in pixel size by almost 30%, since we are talking about matrices of the most modern 960H format with an increased number of pixels up to 976x582 for the PAL standard. Sony offers a range of Effio signal processors to process this signal.

Added IR component
One of the effective methods for increasing the integral sensitivity is the extension of the spectral characteristics of the sensitivity to the infrared range. This is especially true for the Ex-view matrix. Adding an IR component distorts the transfer of the relative brightness of colors somewhat, but this is not critical for the black and white version. The only problem arises with color reproduction in day / night cameras with constant IR sensitivity, that is, without a mechanical IR filter.

The development of this technology in the Ex-view HAD II (ICX658AKA) models in comparison with the previous version (ICX258AK) provides an increase in the integral sensitivity by only 0.8 dB (from 1100 to 1200 mV) with a simultaneous increase in sensitivity at a wavelength of 950 nm by 4, 5 dB. In fig. 1 shows the characteristics of the spectral sensitivity of these matrices, and in Fig. 2 - the ratio of their integral sensitivity.

Optical innovation
Another method of increasing CCD sensitivity is to increase the efficiency of pixel microlenses, light-sensitive area and optimize color filters. In fig. Figure 3 shows the Super HAD and Super HAD II array, showing the enlargement of the lens area and light-sensitive area of \u200b\u200bthe latest modification.

Additionally, Super HAD II matrices significantly increase the transmission of light filters and their resistance to fading. In addition, the transmission in the shortwave region of the spectrum (cyan) is expanded, which improves color reproduction and white balance.

In fig. 4 shows the spectral characteristics of the sensitivity of the Sony 1/3 "Super HAD (ICX229AK) and Super HAD II (ICX649AKA) matrices.

CCD: unique sensitivity

Taken together, the above measures have achieved significant results in improving the performance of CCD.

It is not possible to compare the characteristics of modern models with earlier versions, since then color matrices of widespread use, even of a typical high resolution, were not produced. In turn, now standard definition black-and-white matrices are not produced using the latest technologies Ex-view HAD II and Super HAD II.

In any case, in terms of sensitivity, CCDs are still an unattainable benchmark for CMOS, so they are still widely used with the exception of megapixel versions, which are very expensive and are mainly used for special tasks.

CMOS: pros and cons

CMOS sensors were invented in the late 1970s, but production began only in the 1990s due to technological problems. And at once their main advantages and disadvantages were outlined, which remain relevant even now.

The advantages include greater integration and efficiency of the sensor, wider dynamic range, ease of production and lower cost, especially for megapixel options.

On the other hand, CMOS sensors have a lower sensitivity due to, other things being equal, large losses in filters of the RGB structure, a smaller useful area of \u200b\u200bthe photosensitive element. As a result of the multitude of transition elements, including amplifiers in the path of each pixel, it is much more difficult to ensure the uniformity of the parameters of all sensitive elements in comparison with CCD. But advances in technology have brought CMOS sensitivity closer to the best CCD samples, especially in megapixel versions.

Early proponents of CMOS argued that these structures would be much cheaper because they could be produced on the same hardware and technologies as the memory and logic chips. In many ways, this assumption was confirmed, but not completely, since the improvement of technology led to an almost identical production process in terms of complexity, as for the CCD.

With the expansion of the range of consumers beyond standard television, matrix resolution has grown continuously. These are household camcorders, electronic cameras and cameras built into communications. By the way, for mobile devices the issue of efficiency is quite important, and here the CMOS sensor has no competitors. For example, since the mid-1990s. the resolution of matrices has grown by 1–2 million elements annually and now reaches 10–12 Mpx. Moreover, the demand for CMOS sensors has become dominant and today exceeds 100 million units.

CMOS: improved sensitivity

The first samples of surveillance cameras in the late 1990s - early 2000s with CMOS sensors had a resolution of 352x288 pixels and a sensitivity of about 1 lux even for the black-and-white version. The color versions of the standard definition had a sensitivity of about 7-10 lux.

What suppliers offer
At present, the sensitivity of CMOS matrices has certainly increased, but for typical versions of a color image it does not exceed values \u200b\u200bof the order of several lux at reasonable F values \u200b\u200bof the lens number (1.2–1.4). This is confirmed by the data of the technical characteristics of IP-video surveillance brands, which use CMOS sensors with progressive scan. Manufacturers who claim a sensitivity of about tenths of a lux usually specify that this is data for a lower frame rate, accumulation mode, or at least enabled and sufficiently deep AGC (AGC). Moreover, for some manufacturers of IP cameras, the maximum AGC reaches a mind-boggling value of -120 dB (1 million times). It is hoped that the sensitivity for this case, in the minds of manufacturers, assumes a decent signal-to-noise ratio, allowing you to see more than just snow on the screen.

Innovation improves video quality
In an effort to improve the characteristics of CMOS sensors, Sony has proposed a number of new technologies that provide a practical comparison of CMOS sensors with CCDs in sensitivity, signal-to-noise ratio in megapixel versions.

The new technology for the production of Exmor matrices is based on changing the direction of incidence of the luminous flux on the matrix. In a typical architecture, light strikes the front surface of the silicon wafer through and past the conductors of the matrix circuit. Light is scattered and blocked by these elements. In the new version, light enters the back of the silicon wafer. This led to a significant increase in the sensitivity and noise reduction of the CMOS matrix. In fig. 5 illustrates the difference between the structures of a typical matrix and an Exmor matrix shown in section.

Photo 1 shows images of a test object obtained at 100 lux illumination (F4.0 and 1/30 s) with a CCD (frontal illumination) and CMOS Exmor camera with the same format and resolution of 10 Mpx. Obviously, the CMOS camera image is at least as good as the CCD image.

Another way to improve the sensitivity of CMOS sensors is to abandon the rectangular arrangement of pixels with a line-by-line shift of red and blue elements. In this case, in the construction of one resolution element, two green pixels are used - blue and red from different lines. Instead, it proposes a diagonal arrangement of elements using six adjacent green elements to construct one resolution element. This technology is called ClearVid CMOS. A more powerful image signal processor is assumed for processing. The difference in the structures of the arrangement of colored elements is illustrated in Fig. 6.

The information is read out by a high-speed parallel analog-to-digital converter. At the same time, the frame rate of progressive scan can reach 180 and even 240 fps. Parallel acquisition of information eliminates the diagonal frame shift common for CMOS cameras with sequential exposure and signal readout, the so-called Rolling Shutter effect - when the characteristic blur of fast moving objects is completely absent.

Photo 2 shows images of a rotating fan taken by a CMOS camera at 45 and 180 frames per second.

Full competition

We have cited Sony technologies as examples. Naturally, CMOS-matrices, like CCDs, are produced by other companies, although not on such a scale and not so well-known. In any case, everyone, one way or another, goes roughly the same way and uses similar technical solutions.

In particular, the well-known technology of Panasonic Live-MOS matrices also significantly improves the characteristics of CMOS matrices and, naturally, by similar methods. In Panasonic matrices, the distance from the photodiode to the microlens is reduced. Simplified signal transmission from the photodiode surface. The number of control signals was reduced from 3 (standard CMOS) to 2 (as in CCD), which increased the photosensitive area of \u200b\u200bthe pixel. A low noise photodiode amplifier is used. A thinner sensor layer structure is used. Reduced supply voltage reduces noise and matrix heating.

It can be stated that megapixel CMOS matrices can already successfully compete with CCDs not only in price, but also in such problematic characteristics for this technology as sensitivity and noise level. However, in traditional CCTV television formats, CCD matrices remain unrivaled.

The matrix is \u200b\u200bthe foundation of any photo or video device. It determines the quality and size of the resulting image. Today in the manufacture of matrices, two different technological principles are used - CCD and CMOS. Very often you can hear the question: "Which matrix to choose: CCD or CMOS?" Among fans of photo and video equipment, there are heated debates about this. In this article, we will review these two types and try to figure out which matrix is \u200b\u200bbetter - CCD or CMOS.

general information

Matrices are designed to digitize the parameters of light rays on their surface. It is not possible to speak about a clear advantage of one of the technologies. Comparisons can be made according to specific parameters and a leader can be identified in one aspect or another. As for the preferences of users, often the main criterion for them is the cost of the product, even if it is inferior in quality or technical characteristics to its competitor.

So let's see what both types of devices are. A CCD-matrix is \u200b\u200ba microcircuit that consists of light-sensitive photodiodes; it is silicon-based. The peculiarity of its work lies in the principle of operation of a charge-coupled device. CMOS-matrix is \u200b\u200ba device created on the basis of a semiconductor having an insulated gate with channels of different conductivity.

Principle of operation

Let's move on to identifying the differences that will help determine the choice: which is better - a CMOS matrix or a CCD? The main difference between these two technologies is how they work. CCD devices convert the charge from the pixels into an electrical potential, which is amplified outside the photosensitive sensors. The result is an analog image. After that, the entire picture is digitized in the ADC. That is, the device consists of two parts - the matrix itself and the transducer. CMOS technology is characterized by the fact that it digitizes each pixel separately. The output is a ready-made digital picture. That is, the electric charge in the matrix pixel is accumulated in the capacitor, from which the electric potential is removed. It is transmitted to an analog amplifier (built directly into the pixel), after which it is digitized in a converter.

Should you choose CCD or CMOS?

One of the important parameters that determine the choice between these technologies is the number of matrix amplifiers. CMOS devices have a larger number of these devices (at each point), therefore, when the signal passes through, the picture quality decreases slightly. Therefore, CCDs are used to create images with a high degree of detail, for example, in medical, research, and industrial purposes. But CMOS technologies are used mainly in household appliances: webcams, smartphones, tablets, laptops, etc.

The next parameter that determines which type is better - CCD or CMOS - is the density of the photodiodes. The higher it is, the fewer photons "will be wasted", respectively, the image will be better. In this parameter, CCD matrices bypass their competitors, since they offer a layout that does not have such gaps, while CMOS has them (transistors are located in them).

Nevertheless, when the user faces a choice: which one - CMOS or CCD - to buy, the main parameter pops up - the price of the device. CCD technology is significantly more expensive than its competitor and more energy intensive. Therefore, it is impractical to install them where the image of average quality is sufficient.

CMOS sensor

CMOS arrays use insulated-gate field-effect transistors with channels of different conductivity.

History

In the late 1960s. many researchers have noted that CMOS structures are sensitive to light. However, CCD devices provided such a higher light sensitivity and image quality that CMOS sensors did not receive any noticeable development.

In the early 1990s, the characteristics of CMOS sensors, as well as manufacturing technology, were significantly improved. Advances in submicron photolithography have made it possible to use thinner compounds in CMOS sensors. This led to an increase in light sensitivity due to a larger percentage of the irradiated area of \u200b\u200bthe matrix.

The revolution in CMOS sensor technology came when NASA's Jet Propulsion Laboratory (JPL) successfully implemented Active Pixel Sensors (APS) - active pixel sensors . Theoretical studies were carried out several decades ago, but the practical use of an active sensor was postponed until 1993. APS adds a transistor amplifier to each pixel for readout, which makes it possible to convert charge into voltage directly in the pixel. It also provided random access to photodetectors, similar to that implemented in RAM chips.

As a result, by 2008 CMOS had become practically an alternative to CCDs.

At the MWC in Barcelona this year, Samsung demonstrated a new type of CMOS sensors that are targeted at smartphones.

Principle of operation

• Reset signal is given before shooting
• During exposure, the charge is accumulated by the photodiode
• In the process of reading, the voltage value on the capacitor is sampled

Benefits

• The main advantage of CMOS technology is its low static power consumption. This makes it possible to use such matrices as part of non-volatile devices, for example, in motion sensors and surveillance systems that are in the "sleep" or "waiting for an event" mode most of the time.
• An important advantage of the CMOS matrix is \u200b\u200bthe unity of the technology with the rest, digital elements of the equipment. This leads to the possibility of combining analog, digital and processing parts on a single crystal (CMOS technology, being primarily a processor technology, implies not only the "capture" of light, but also the process of converting, processing, cleaning signals not only actually captured, but and third-party components of the electronic equipment), which served as the basis for miniaturization of cameras for a variety of equipment and reducing their cost due to the rejection of additional processor chips.
• The random access mechanism can read selected groups of pixels. This operation is called windowing readout. Cropping allows you to reduce the size of the captured image and potentially increase the read speed compared to CCD sensors, since in the latter, all information must be unloaded for further processing. It becomes possible to use the same matrix in fundamentally different modes. In particular, by quickly reading only a small part of the pixels, it is possible to provide a high-quality live image viewing mode on the screen built into the device with a relatively small number of pixels. You can scan only part of the frame and apply it to full screen display. Thus, to get the opportunity to high-quality manual focusing. It is possible to conduct reportage high-speed shooting with a smaller frame size and resolution.
• In addition to the amplifier within the pixel, amplification circuits can be placed anywhere along the signal path. This allows you to create amplification stages and increase sensitivity in low light conditions. The ability to change the gain for each color improves, in particular, the white balance.
• Low cost of production in comparison with CCD matrices, especially with large matrix sizes.

disadvantages

• The cell photodiode occupies a significantly smaller area of \u200b\u200bthe matrix element compared to a full-frame transfer CCD. Therefore, early CMOS sensors had significantly lower light sensitivity than CCDs. But in 2007, Sony launched a new line of next-generation CMOS video and cameras with Exmor technology, which was previously only used for CMOS sensors in specific optical devices such as electron telescopes. In these matrices, the electronic "binding" of the pixel, which prevents photons from hitting the light-sensitive element, was moved from the upper to the lower matrix layer, which made it possible to increase both the physical size of the pixel with the same geometric dimensions of the matrix and the accessibility of the elements to light, which, accordingly, increased light sensitivity of each pixel and the matrix as a whole. For the first time, CMOS matrices were compared with CCD matrices in terms of light sensitivity, but they turned out to be more energy-efficient and devoid of the main drawback of CCD technology - the "fear" of point light. In 2009, Sony improved EXMOR CMOS sensors with "Backlight illumination" technology. The idea of \u200b\u200bthe technology is simple and fully corresponds to the name.
• The photodiode of the matrix cell has a relatively small size, while the value of the resulting output voltage depends not only on the parameters of the photodiode itself, but also on the properties of each pixel element. Thus, each pixel of the matrix has its own characteristic curve, and the problem of scatter arises

Camera, features, advantages and disadvantages of such matrices.

To the merits CCDs can be attributed:

• High utilization of pixel area (close to 100%);
• relatively low;
• very high efficiency;
• big enough .

The disadvantages CCDs relate:

• high energy intensity;
• rather complicated process of reading information;
• expensive production.

Modern digital cameras use not only CCD-based matrices, but also CMOS sensors, the share of cameras equipped with such matrices is constantly growing.

CMOS-matrix of the camera.

Back in the late 60s of the last century, scientists knew the property of CMOS structures to perceive light. However, CCD structures provided much higher sensitivity to light and high image quality. This is why CMOS matrices have not been widely adopted. Early 90s characteristics CMOS matrices and their production has been greatly improved, leading to wider adoption of these dies. Groundbreaking discoveries were made in the Jet Propulsion Laboratory (JPL NASA), where Active Pixel Sensors (APS) were created. The bottom line was that a transistor signal amplifier was added to each one, which made it possible to convert charge into voltage directly in the pixel itself. Thanks to this, random access to individual pixels became possible, in principle similar to RAM circuits.

As a result, by 2008, CMOS matrices became an alternative to CCD matrices.

CMOS-matrix (complementary metal-oxide-semiconductor structure), in English transcription - CMOS (Complementary metal oxide semiconductor), in principle, is similar to a CCD matrix. Just like in a CCD, electrons are produced under the influence of light.

CMOS cells are insulated-gate field-effect transistors and have channels of different conductivity.

Unlike a CCD element, each cell CMOS sensors additionally has electronic devices, called pixel straps, that convert charge into voltage directly in the cell.

Figure 1 shows the equivalent circuit for a CMOS element.

Fig. 1. Equivalent electrical circuit of a CMOS element.

1 - LED. 2 - electronic shutter. 3 - capacitor that accumulates the charge from the photodiode. 4 - signal amplifier. 5 - line reading bus. 6 - bus through which the signal is transmitted to the processor. 7 - line for supplying a reset signal.

The principle of operation of the above circuit:

before the image is taken, a reset signal is sent along line 7;

when light is applied to the photodiode, a charge is created in it in proportion to the intensity of the light flux, which charges the capacitor;

the signal from the element is read out by discharging the capacitor, the resulting current is transmitted to the amplifier and then to the processing circuit.

Synchronization of the matrix is \u200b\u200bcarried out through the address buses of columns and rows.

Thanks to such a scheme, it becomes possible to read the charge immediately from a group of pixels (and not sequentially cell by cell, as in a CCD matrix) or even selectively from individual pixels. In such a matrix, there is no need for column and row shift registers, which greatly speeds up the process of reading information from the matrix,. The power consumption of the matrix is \u200b\u200balso significantly reduced.

Progress in the development of technologies, in particular, the production of high-quality silicon wafers and the improvement of the amplifier circuit of the CMOS element, led to the fact that the latter reached the quality of the resulting image almost at the same level as the CCD element.

CMOS advantages:

First of all, power consumption has been significantly reduced, due to the fact that in a CMOS matrix the information processing chain is not as long as in a CCD matrix, especially in a low power consumption, the CMOS matrix differs in static mode.

The CMOS matrix cell circuit allows it to be integrated directly with an analog-to-digital converter and even with a processor. This makes it possible to combine in one crystal both an analog circuit and a digital and processing one. Thanks to this, it became possible to further miniaturize digital cameras, reduce their cost due to the absence of the need for additional processor chips.

Random access to CMOS cells allows individual pixel groups to be read. This feature is called cropped reading, that is, reading only a part of the entire frame, in contrast to the CCD matrix, where it is necessary to unload the entire matrix for information processing. This allows only a fraction of the information to be displayed on the camera's built-in display with a relatively small number of pixels to ensure quick viewing of the image. This will be enough for viewing, you can control the focusing accuracy, etc.

In addition, for a higher speed of sequential shooting, you can conduct it with a smaller frame size and lower resolution.

Another advantage of the CMOS matrix is \u200b\u200bthe ability to add amplifying stages to the amplifier inside the CMOS element, thereby significantly increasing the sensitivity of the matrix. And the ability to adjust the gain for each color can improve.

Manufacturing CMOS arrays is simpler and cheaper than CCDs, it can be mastered by almost any plant that produces microelectronics. This is especially true when producing large matrices.

Disadvantages of a CMOS sensor:

The disadvantages of a CMOS matrix in comparison with a CCD matrix include, first of all, a decrease in the photosensitive part of the element due to the presence of an electronic strapping around the pixel. That is why, in the beginning, CMOS sensors had significantly lower sensitivity than CCDs. This changed with the development and marketing by Sony in 2007 of CMOS sensors, manufactured using EXMOR technology, previously used for specific devices such as electron telescopes. The size of the photosensitive part of the pixel was increased by moving the electronic strapping to the lower layer of the element, where it did not interfere with the ingress of light. This has led to an increase in the sensitivity of each pixel and the entire matrix.

In each of the elements of the CMOS matrix there are also electronic elements, which, by the properties of electronic circuits, have their own noise, and this noise is added to the noise of the photosensitive element itself. Moreover, for each pixel the level of this noise is different.

The magnitude of the signal received from each pixel depends not only on the characteristics of the photodiode itself, but also on the properties of each element of the electronic strapping of the pixel. Hence, each CMOS element has its own