Among analog cameras and standard definition network cameras, CCD is the most widely used and has occupied a leading position in the market for a long time. CCD is characterized by high sensitivity, but slow response, which is not suitable for high-resolution progressive scanning mode used by high-definition surveillance cameras. Therefore, after entering the era of high-definition surveillance, CMOS is gradually recognized by people, and CMOS photosensitive devices are widely used in high-definition surveillance cameras.
The main advantage of CMOS used in CCD is very power saving. Unlike CCD composed of diodes, CMOS circuits have almost no static power consumption. This makes the power consumption of CMOS only about 1/3 of that of ordinary CCD. The important problem of CMOS is that when dealing with fast-changing images, it is overheated because of too frequent current conversion, so it is not a big problem to suppress dark current well, and it is very easy to appear noise if it is not suppressed well.
720P and 1080P back-illuminated CMOS devices have been developed, and their sensitivity performance is close to that of CCD. Compared with the surface-illuminated CMOS sensor, the back-illuminated CMOS sensor has great advantages in sensitivity (S/N), which significantly improves the shooting effect in low illumination, so the noise can be greatly reduced when shooting in low illumination.
Although the megapixel camera products based on CMOS technology have some shortcomings in low illumination environment and signal noise processing, it will not fundamentally affect its application prospects. In addition, relevant international enterprises are stepping up efforts to solve these two problems. I believe that in the near future, the effect of CMOS will be closer and closer to CCD, and the price of CMOS equipment will be lower than CCD equipment.
It has become an indisputable fact that CMOS is more used in security industry than CCD. Although the resolution of CCD sensor with the same size is better than that of CMOS sensor, if the size limit is not considered, the advantage of CMOS in quantity rate can effectively overcome the difficulty of manufacturing large-size photosensitive elements, so that CMOS will have an advantage in higher resolution. In addition, the response speed of CMOS is faster than that of CCD, which is more suitable for high-definition monitoring of big data. Compared with CCD, CMOS has the advantages of small size, lower power consumption than110 and lower price than CCD 1/3.
Compared with CCD products, CMOS is a standard process, which can make use of existing semiconductor equipment without additional investment equipment, and the quality can be improved with the improvement of semiconductor technology. At the same time, there are many CMOS production lines in global fabs, which will also help to reduce costs in mass production in the future. In addition, the biggest advantage of CMOS sensor is that it has high system integration. Theoretically, all the functions required by the image sensor, such as vertical displacement, horizontal displacement register, timing control, CDS, ADC…, etc. , can be integrated on one chip, and even all chips including back-end chips and Flash RAM can be integrated into a system on chip, thus reducing the production cost of the whole machine.
Because of this, there are more manufacturers investing in R&D, including more than 30 in the United States, 7 in Europe, about 8 in Japan, 8 in Korea 1 and 8 in Taiwan Province Province. The leading manufacturer in the world is Agilent (Hewlett-Packard), with a market share of 565,438+0%, ST (Very Large Scale Integrated Circuit Vision) of 65,438+06%, Omni Vision of 65,438+03%, Hyundai of 8% and Photobit of about 5%, with a total market share of 93%.
According to In-Stat statistics, by 2004, the global sales of CMOS sensors are expected to exceed 654.38+0.8 billion US dollars, and CMOS will grow rapidly with a compound annual growth rate of 62%, gradually eroding the application fields of CCD devices. Especially in the rapidly developing mobile phone application field of 20 13, the camera module based on CMOS image sensor will occupy more than 80% of its application market. CMOS image sensor belongs to the new product market, and its market share is not as constant as that of mature industries. For example, in CMOS market, Agilent, OmniVision, STM, and Hyundai are ranked according to the proportion of shipments, with market shares of 24%, 22%, 14%, and 14% respectively, among which STM. However, after only one year of market competition, Agilent and OmniVision still ranked first and second, and their market share increased to 37.7% and 30.8% respectively, while STM ranked fourth, and its market share dropped sharply to 4.8%. As for Hyundai, its market share dropped sharply to 2. 1%. It is worth mentioning that Photobi grew rapidly in the global market in 2000. The shipments of these three manufacturers accounted for 82.2% of the global shipments. From this analysis, we can see that the concentration of manufacturers in this industry is quite dense, so observing the dynamics and development of the above three manufacturers, we can see the future development direction of Xu industry and technology.
Agilent's main products are the second generation CIF(352*288)HDCS- 1020 and the second generation VGA(640*480)HDCS-2020, which are mainly used in digital cameras, mobile phones, PDA, PC cameras and other emerging information appliances. In addition, another successful strategy of Agilent in 2000 was to form a strategic alliance with Logitech and Microsoft to enter the field of optical mouse products. But this is a very low-grade CMOS product, and it is not used to shoot images, so this figure is not included in the global statistics of image sensors, but it shows Agilent's planning intention to enter the optical components based on CMOS technology.
The main products of OmniVision include: CIF (352x288), VGA(640 x 480), SVGA(800 x 600) and SXGA( 1280 x 1024). The CMOS image sensor with 6.5438+0.3 million pixels developed by Omnivision is being widely used in digital cameras. The industry generally believes that one million pixels is the watershed of using CMOS and CCD. The successful entry of CMOS into this market shows that the development of CMOS technology has penetrated into the market and may replace CCD as the application of low-end image products in the future. The CMOS sensor of CIF(352 x 288) level developed by OmniVision in May 20065438+0 is characterized by low power consumption. Its target market is mobile phones, and its product development strategy coincides with that of major research institutions. In the mobile phone market, the camera module of CMOS module has become the product with the largest number of mobile communication applications.
Photobit was a great success in 2000. In 200 1 year, Photobit took the lead in developing the CMOS image sensor of PB-0330 product model. This product uses a single-chip logic-digital converter, which is the second generation of 1/4-inch VGA(640 x 480), and also introduces the CMOS image sense of Pb-01/product model. Photobit launched these two products mainly for digital cameras and PC cameras, which are different from OmniVision CIF(352 x 288) in the mobile phone market. It has introduced two image sensors with different resolutions, CIF(352 x 288) and VGA(640 x 480), and its marketing scope is intended to cover the low-end and high-end markets. 20 13, the industry has developed a new technology of CMOS image sensor-C3D. The biggest feature of C3D technology is the uniformity of pixel reaction. C3D technology redefines the performance of imager (including the overall performance of the system), and improves the standard performance of CMOS image sensor in terms of uniformity and dark current.
At the beginning of 20 14, American company Foveon publicly demonstrated its newly developed Foveon X3 technology, which immediately attracted great attention from the industry. Foveon X3 is the first image sensor array in the world that can capture all colors on one pixel. Traditional photoelectric coupling devices can only sense light intensity, but can't sense color information. It needs to perceive color information through a color filter, which we call Bayer color filter. On the other hand, Foveon X3 senses color at different depths of a pixel, with the outermost layer sensing blue, the second layer sensing green and the third layer sensing red. It is based on the absorption effect of silicon on light with different wavelengths to realize that a pixel perceives all color information. CMOS image sensors have adopted this technology, and its application product is "Sigma SD9" digital camera.
This innovative technology can provide clearer images and better colors. Compared with the previous image sensor, X3 is the first sensor to detect color through the built-in silicon photoelectric sensor. Of Foveon's technology has made a major breakthrough in the traditional semiconductor photosensitive technology, and also has the effect of subverting the traditional technology. I believe Foveon X3 will have a bright future.
In terms of high-resolution pixel products, a few days ago, Taiwan Province Ruishi Technology has led the industry in mass production of 265,438+megapixel CMOS image sensors. American companies cooperated with Taiwan Province Optical Lens Factory to launch this module of CMOS sensor combined lens in the third quarter, and CMOS applications have begun to be applied in 2 megapixel digital camera products.
CMOS linear array image sensor DLIS-2K- the fastest single-port reconstruction linear image sensor in the world. Measurement range: 200nm~ 1 100nm output signal: digital DLIS-2K linear array image sensor includes four rows of pixels, each row has 208 1 optical pixels and 16 black pixels. Among them, three rows are square pixels of 4×4 microns, and the other row is rectangular pixels of 4×32 microns. Using correlated multiple sampling (CMS) method, its equivalent sensitivity can reach 160 V/lux-s ... each line can control the exposure and output at will. In addition, the background sampling can use the traditional correlated double sampling (CDS) value, or be set as the ambient light value controlled by the user. The sensor is controlled by a 3-wire serial port, which integrates our patented technology, high-speed distributed 8-to 1 1 bit analog-to-digital converter (D/AD), XtremEIX and active column sensor technology to maximize the application function. Further reading: 2k series DLIS configurable line scanning CMOS image sensor. DLIS-2K sensors all use the advanced photodiode (APD) pixel technology of the building and the imaging pixel IP architecture patented by Panavishen. These reconfigured linear image sensors provide high performance at low cost, and combine high sensitivity, high speed and multifunction, which can solve many applications in consumer electronics, industry, automobile and technology markets. According to the analysis, the valuation of global industrial companies is the world image sensor market,11700 million dollars, rising to 20 12. Generally speaking, image sensors have been extended to industrial and commercial fields, such as biometric technology, machine vision, broadcasting, movie cameras and medicine, such as video cameras, security and computer cameras, portable communication devices and consumer electronics applications. In the automobile industry, there is a demand for increasing the angular rate, occupying seats, cruise control sensors, lane departure systems and rearview cameras. The imager of DLIS-2K is a four-wire sensor with 1 1 bit analog-to-digital conversion, high dynamic range and improved sensitivity related to multiple sampling (CMS). The sensor can be used in spectroscopy, bar code, touch screen, optical character recognition, machine vision, measurement and other applications. The progress of these patented technologies makes the products flexible in image acquisition and reading, including ambient light subtraction, oversampling, lossless reading mode, different integration, automatic threshold and pixel reading of 120MHz, which is unprecedented in high-resolution mode. DLIS sensor environment combination 12 bit digitization and automatic threshold light subtraction. This provides a simple binary output chip that allows the centroid removal of barcodes, touch screens or any application that needs to find a location or many parts of the system. Users can also input analog signals, that is, applications may need to be digitized. Operation modes can be mixed or matched, and there are four possible pixel combinations for many different applications, making this series the best solution. "Our goal is to solve the problem at a very competitive price and expand the market of bar codes and programmable image sensor touch screens. Tower's CMOS image sensor technology and manufacturing capability are world-class, and the close interaction between our design team and tower engineers helps us achieve rapid production, "said Jeffrey Zarnowski, chief technology officer of Panavishen Imaging Company. "We are very satisfied with Panavishen's linear image sensor series, because these products will greatly promote the production capacity of countless devices in various markets. By combining our advanced photo diode (APD) pixel technology and pixel IP with Panavishen's patented imaging architecture, we have realized the imaging function that could not be achieved by linear imaging before, "said Dr. Avisrum, vice president and general manager of Tower Semiconductor's professional business department. The tower's 0. 18 micron technology realizes on-chip and bit-selectable analog-to-digital converter and higher data transmission rate than previous products. The APD process of the tower shows improvement, and the IP of the pixel is more sensitive than the standard photodiode with high charge transfer characteristics. The technology of tower and the combination of Panavishen image and architecture make the pixel with 4 × 32 micron sensitivity exceed 100 volt/lux second.