Basic introduction
Other digital signal processing circuits can also be integrated on the CMOS image sensor chip, such as AD converter, automatic exposure control, non-uniform compensation, white balance processing, black level control, Gamma correction, etc., in order to perform fast calculations, even DSP devices with programmable functions can be integrated with CMOS devices to form a single-chip digital camera and image processing system.
In 1963, Morrison published a computable sensor, which is a structure that can use the light guide effect to determine the position of the light spot, which became the beginning of the development of CMOS image sensors. In 1995, the low-noise CMOS active pixel sensor single-chip digital camera was successful.
CMOS image sensor has the following advantages: 1) Random window reading capability. The random window reading operation is an aspect of the CMOS image sensor that is better than the CCD in function, and it is also called the selection of the region of interest. In addition, the highly integrated characteristics of CMOS image sensors make it easy to open multiple tracking windows at the same time. 2) Ability to resist radiation. In general, the potential anti-radiation performance of CMOS image sensors has an important enhancement over CCD performance. 3) The complexity and reliability of the system. Using CMOS image sensor can greatly simplify the system hardware structure. 4) Non-destructive data readout method. 5), Optimized exposure control. It is worth noting that due to the integration of multiple functional transistors in the pixel structure, CMOS image sensors also have several shortcomings, mainly two indicators of noise and fill rate. In view of the relatively superior performance of CMOS image sensors, CMOS image sensors have been widely used in various fields.
U.S. high-definition high-speed CMOS image sensor
DYNAMAX-11: Panavision Imaging’s new sensor contains global electronic exposure shutter technology, which greatly improves industrial imaging indoors and outdoors Applications. This newly released DYNAMAX-11 image sensor is suitable for industrial imaging fields such as machine vision, security monitoring, intelligent transportation, life science, biomedicine, scientific imaging, high-definition video, TV broadcasting, etc. This newly released DYNAMAX-11 image sensor contains 3.2 million pixels with a pixel size of 5.0 m×5.0 m. DYNAMAX-11 has the following features:
1: High sensitivity, low noise. DYNAMAX-11 can achieve less than 4electronsrms noise in rolling exposure mode, and can achieve less than 8electronsrms noise in global exposure mode.
2: Wide spectral response range, covering from visible light to infrared.
3: DYNAMAX-11 has a fast output capability, which can reach a full-size 3.2M output at 60 frames per second, and an HDTV 1920*1080 output at 72 frames per second.
4: The dynamic range in high dynamic mode can reach 120 decibels.
DYNAMAX-11 adopts CLCC package, which is very convenient for customers' installation, welding and structural design. DYNAMAX-11 is suitable for 3/4 inch optical size. At the same time, DYNAMAX-11 corresponds to the high-definition television format requirements (HDTV, 1080i, 16:9), and also designed a 2/3-inch 2 million pixel optical format (diagonal 11 mm) in the area of interest.
The samples of DYNAMAX-11 color and black and white chips are being provided to PVI customers.
Basic principle
Basic working principle of CMOS image sensor
First of all, the external light irradiates the pixel array, a photoelectric effect occurs, and a corresponding charge is generated in the pixel unit. The row selection logic unit selects the corresponding row pixel unit as required. The image signal in the row pixel unit is transmitted to the corresponding analog signal processing unit and A/D converter through the signal bus of the respective column, and converted into a digital image signal for output. The row selection logic unit can scan the pixel array progressively or interlacedly. The row selection logic unit and the column selection logic unit are used together to realize the window extraction function of the image. The main function of the analog signal processing unit is to amplify the signal and improve the signal-to-noise ratio. In addition, in order to obtain a practical camera with qualified quality, various control circuits must be included in the chip, such as exposure time control, automatic gain control, and so on. In order to make each part of the circuit in the chip operate according to the specified tempo, multiple timing control signals must be used. In order to facilitate the application of the camera, the chip is also required to output some timing signals, such as synchronization signals, line start signals, field start signals, and so on.
The working principle of the pixel array
An intuitive performance index of the image sensor is the ability to reproduce the image. And the pixel array is the key functional module directly related to this index. According to the structure of the pixel array unit, the pixel unit can be divided into passive pixel unit PPS (passivepixelschematic), active pixel unit APS (activepixelschematic) and logarithmic pixel unit. Active pixel unit APS can be divided into photodiode type APS , Raster type APS.
The above various pixel array units have their own characteristics, but they have basically the same working principle. The following first introduces their basic working principles, and then introduces the characteristics of various pixel units. The figure below is a schematic diagram of a single pixel.
(1) First enter the "reset state", then open the gate tube M. The capacitor is charged to V, and the diode is in the reverse state;
(2) Then enter the "sample State". At this time, the gate tube M is closed, and the diode generates a photocurrent under the light to discharge the stored charge on the capacitor. After a fixed time interval, the amount of charge stored on the capacitor C is proportional to the light. An image has been taken into the sensitive element array;
(3) Finally, it enters the "readout state". At this time, open the gate tube M, and read the capacitor C in each pixel one by one. Stored charge voltage.
The passive pixel unit PPS appeared the earliest, and the structure has not changed much since its appearance. The passive pixel unit PPS has a simple structure, high pixel filling rate, and relatively high quantum efficiency, but it has two significant shortcomings. One is that its readout noise is relatively large, and its typical value is 20 electrons, while the typical value of its readout noise for commercial CCD-level technology chips is 20 electrons. Second, as the number of pixels increases, the readout rate increases, so the readout noise becomes larger.
The quantum efficiency of the photodiode type APS is relatively high. Due to the adoption of the new noise elimination technology, the quality of the output graphic signal is much higher than before. The readout noise is generally 75-100 electrons. C3& is suitable for medium and low-end applications.
In the grating type APS structure, the fixed pattern noise is suppressed. Its read noise is 10-20 electrons. But its process is more complicated, and it cannot be regarded as a complete CMOS process strictly speaking. Due to the introduction of the polysilicon cover layer, its quantum efficiency is relatively low, especially for blue light. At present, its overall performance advantage is not very prominent.
Factors affecting performance
3.1 Noise
This is the primary issue that affects the performance of CMOS sensors. This kind of noise includes fixed pattern noise FPN (Fixed pattern noise), dark current noise, thermal noise and so on. The reason for the fixed pattern noise is that the output signals generated by a beam of the same light shining on two different pixels are not exactly the same. Noise is introduced in this way. To deal with fixed pattern noise, double sampling or correlated double sampling techniques can be applied. Specifically, it is a bit like introducing a differential pair when designing an analog amplifier to suppress common mode noise. Double sampling is to first read out the charge integration signal generated by the illumination, temporarily store it, then reset the pixel unit, and then read the output signal of the pixel unit. The image signal is obtained by subtracting the two. Both types of sampling can effectively suppress fixed pattern noise. In addition, correlated double sampling requires a temporary storage unit. As the number of pixels increases, the storage unit also increases.
3.2 Dark current
Physical devices cannot be ideal. Like the subthreshold effect, due to impurities, heat and other reasons, even if no light is irradiated to the pixel, like The element unit also generates charges, and these charges generate dark currents. It is difficult to distinguish between dark current and electric charge generated by light. The dark current is not exactly the same everywhere in the pixel array, and it causes fixed pattern noise. For a pixel unit with an integration function, the fixed pattern noise caused by the dark current is proportional to the integration time. The generation of dark current is also a random process, which is a source of shot noise. Therefore, the dark current generated by the thermal noise element is equal to the square root of the number of dark current electrons in the pixel unit. When a long-term integration unit is used, this type of noise becomes the main factor affecting the quality of the image signal. For dim objects, long-term integration is necessary, and the capacitance of the pixel unit is limited, so it is dark. The accumulation of current electrons limits the maximum time of integration.
In order to reduce the influence of dark current on the image signal, first of all, cooling measures can be taken. However, cooling the chip is far from enough, and the fixed pattern noise generated by the dark current cannot be completely overcome by double sampling. The effective method adopted is to subtract the reference dark current signal from the obtained image signal.
3.3 pixel saturation and overflow blur
Similar to the amplifier, there is an upper input limit due to the limited range of the linear region. For CMOS image sensor chips, it also has an input The upper limit. If the input light signal exceeds this upper limit, the pixel unit will be saturated and cannot perform photoelectric conversion. For a pixel unit with an integral function, the upper limit is determined by the capacity of the optoelectronic integral unit: for a pixel unit without an integral function, the upper limit is determined by the maximum current flowing through the photodiode or transistor. When the input optical signal is saturated, overflow blur occurs. The overflow blur is due to the saturation of the photoelectrons of the pixel unit and flowing out to the neighboring pixel unit. The overflow blur is reflected on the image as a particularly bright area. This is somewhat similar to overexposure in photos. Overflow blur can be overcome by adding an automatic drain tube in the pixel unit, which can effectively discharge the excess charge. However, this only limits the overflow, but cannot make the pixels truly restore the image.
Market status
According to the market research company CahnersIn-statGroup, the imaging products based on CMOS image sensors will reach more than 50% in the next few years. That is to say, CMOS image The sensor will replace the CCD and become the mainstream of the market. It can be seen that the market prospect of CMOS cameras is very broad.
In the next few years, the global sales of CMOS image sensors will increase rapidly, and will impact traditional CCDs in many digital image applications. This is because CMOS image sensor devices have two major advantages: one is that the price is 15% to 25% lower than that of CCD devices; the other is that the chip structure can be easily integrated with other silicon-based components, which can effectively reduce the overall system cost. Although the image quality of CMOS image sensors in the past is worse than that of CCD and the resolution is lower, but after rapid improvement, it has continuously approached the technical level of CCD. This kind of sensor has been widely used in digital cameras, electronic toys, and electronic toys that require lower resolution. In the camera structure of video conferencing and security systems.
The low-resolution digital camera using CMOS image sensor launched by Japan's Nintendo Co., Ltd. has sold 1 million units in the first two months of its launch. Mitsubishi Corporation, Motorola, Hewlett-Packard, Toshiba and Intel also listed this type of product.
Application
Application of CMOS image sensor
1. Digital camera
People have used film cameras for hundreds of years, in the 20th century Since the 1980s, people have used high technology to develop CCD digital cameras that do not require film. Makes a fundamental change to the traditional film camera. The emergence of low-cost FLASHROM that can be written and controlled by electricity, and the advent of low-power, low-cost CMOS cameras. A new situation has been opened for the digital camera. The functional block diagram of the digital camera is shown in the lower right figure.
As can be seen from the figure, the internal device of the digital camera is completely different from the traditional camera. The color CMOS camera takes a picture under the control of the electronic shutter and saves it in DRAM, and then transfers it to Stored in FLASHROM. According to the capacity of FLASHROM and the compression level of image data, the number of photos that can be stored can be determined. If you change the ROM to a PCMCIA card, you can expand the capacity of the digital camera by changing the card, which is like changing a film, transferring the digital image information of the digital camera to the hard disk of the PC for storage, which is very convenient Storage, retrieval, processing, editing and transmission of photos.
2. CMOS digital camera
The USB camera composed of the OV7610 CMOS color digital image chip, the OV511 advanced camera and the USB interface chip launched by the American OmniVison company, its resolution Up to 640x480, suitable for video systems transmitted via Universal Serial Bus. The introduction of the OV511 advanced camera allows the PC to obtain a large amount of video information in a more real-time manner. The compression ratio of the compression chip can reach 7:1, thus ensuring the rapid image transmission from the image sensor to the PC. For the CIF image format, the OV511 type can support a transmission rate of up to 30 frames per second, reducing the image jitter that usually occurs in low-bandwidth applications. As a high-performance USB interface controller, OV511 has sufficient flexibility and is suitable for occasions including video conferencing, video e-mail, computer multimedia and security monitoring.
3. Application in other fields
CMOS image sensor is a multifunctional sensor. Because it has the performance of CCD image sensor, it can enter the application field of CCD, but it also It has its own unique advantages, so it has opened up many new application areas. In addition to the main applications described above, CMOS image sensors can also be applied to digital still cameras and small medical cameras. For example, a cardiac surgeon can install a small "silicon eye" on the patient's chest to monitor the effect of the operation after the operation. CCD is difficult to achieve this application.
4. Applied to the X-ray machine market
In the dental X-ray machine market, use