DSP chip plays an important role in camera, which greatly improves the technology of traditional camera. In the application of camera, the signal generated by optical imaging chip inevitably has many defects, such as weak image detail signal, uneven brightness, unnatural color and so on. The signal must be amplified and corrected by DSP. As the core component of the camera, the performance of DSP processing chip directly determines the quality of the camera monitoring picture.
DSP (digital signal processing) is the use of digital computer or special digital signal processing equipment, with the method of numerical calculation for signal acquisition, transformation, synthesis, evaluation, recognition and other processing, in order to achieve the purpose of extracting useful information, convenient for application.
DSP chip is a special microprocessor. According to the mathematical model and algorithm of digital signal processing theory, a special digital signal microprocessor chip is designed. All computing programs are hardened, and other devices needed by digital filter are integrated and hardened, such as adder, memory, controller, I / O interface, and even other types of external devices. Many works that can not be done in analog signal processor can be done in digital processing, such as two-dimensional digital filtering, digital dynamic image detection, digital background light compensation, skin color contour correction, detail compensation frequency adjustment, accurate color matrix, accurate correction, automatic focusing, etc. Therefore, with the DSP camera, the image quality can be greatly improved.
The function of camera DSP chip
Image quality is the most basic and important requirement of monitoring. The monitoring scenes are very different, which requires that the monitoring camera can adapt to the complex light of various scenes and obtain excellent image quality. Image quality contains a lot of content, mainly including the following points:
1. Color restoration. Whether the camera can truly restore the image color of the actual scene, whether the white balance is accurate, whether there is warm or cold.
2. Dynamic effects. Whether the edge of the moving object is clear and whether there is trailing.
3. Low illumination performance. In the low light environment, whether the camera can get a bright image, whether the sensitivity is excellent.
4. Wide dynamic. In the scene with strong light dark contrast, whether the camera can take care of both bright and dark areas, and whether the dynamic range is large enough.
5. Focus accuracy. Whether the camera can focus accurately under various lighting conditions without losing focus or hesitating repeatedly.
The quality of camera image largely depends on the post-processing of DSP. In the application of camera, the signal generated by optical imaging chip inevitably has many defects, such as weak image detail signal, uneven brightness, unnatural color and so on. The signal must be amplified and corrected by DSP signal processor.
In addition to the image signal amplification and correction, the camera DSP chip also has a good performance in the realization of the special functions of the camera. Such as auto focus, wide dynamic, digital slow shutter, color black-and-white conversion, area shielding, digital noise reduction, white balance and other functions are derived from DSP digital signal processor. Therefore, the breakthrough in the definition of surveillance cameras, especially analog cameras, largely depends on DSP, which makes analog cameras still occupy a large share in some fields.
DSP chip technology upgrade progress
Increasingly fierce competition
As the core component of the camera, the performance of DSP processing chip directly determines the quality of the camera monitoring picture. At present, domestic camera manufacturers do not have the strength to research and develop core chip processing technology. To a large extent, they rely on the purchase of foreign chip products to assemble their own cameras. Therefore, the leading brand image and huge market demand have prompted many well-known manufacturers to enhance their own core R & D strength, making the chip war intensified and upgraded on the basis of previous years.
In recent years, the chip R & D investment of major manufacturers has been continuous. For example, Sony launched the latest dyna view ex wide dynamic technology chip, and integrated depa second generation intelligent audio and video alarm technology, Visibility Enhancement Technology and super dynamic noise reduction technology on the chip, which greatly improved the chip technology, thus enhancing the image effect of the camera. Samsung Electronics launched the A1 HD smart chip, which has intelligent recognition technology, so that the camera can identify suspicious people at the front end, which is more humanized. Samsung optoelectronics has launched the fourth generation of SV-IV high-performance DSP image engine. The chip integrates precision image analysis technology, ssnriii and newly upgraded 2D / 3D image noise reduction technology, which greatly improves the image quality of the camera. Panasonic has launched analog and digital cameras with the fifth generation super wide dynamic technology (SD5). It is not only limited to the traditional shutter and signal processing technology to make the overall image performance, but also integrates three core technologies: Super Dynamic + automatic dark area compensation + intelligent mobile detection. LG’s xdi-sdsp chip was successfully put into production in 2009. It integrates the past DSP, OPD, OSD, TG and other chips into one. The single-chip solution has greatly improved the picture performance and stability. LG’s latest xdi-s2 DSP technology has made significant breakthroughs in horizontal definition, color restoration, automatic white balance and low illumination noise reduction.
Ding Quan, technical engineer of security division of Lejin Electronics (China) Co., Ltd., said that at present, most manufacturers have increased the internal algorithms of DSP, mainly in improving its performance, clarity, color, low illumination and other aspects that everyone pays attention to, making the image effect of cameras on the market increasingly fine; At the same time of improving the performance, each company has also invested a lot of R & D efforts in its stability and power consumption, to stay in the market with more excellent performance and excellent quality. In the face of increasingly fierce market competition, manufacturers’ requirements for other aspects are gradually increasing. In the past, a DSP solution can fully meet the market in a few years, but now the market demand is increasing, it is necessary to make the DSP solution constantly updated, from the continuous improvement of clarity to the improvement of various functions. In this way, the matching design of multiple schemes affects the image effect of the whole scheme. At the same time, the performance and function of DSP have been comprehensively developed in the face of the continuous shortening of product update speed. In order to achieve the functions of menu, white balance and color restoration, traditional cameras need to use a variety of solutions, which brings challenges to the power consumption, compatibility and stability of cameras.
Application of DSP in intelligent camera
With the development of modern science and technology, the video monitoring system is also developing towards intelligence, and the camera for monitoring must also be intelligent. There are two types of DSP cameras: ordinary and intelligent. In the past, DSP intelligent cameras were mainly used to improve image quality. The new detection recognition alarm DSP intelligent camera has more powerful image processing recognition ability and intelligent factors than the ordinary DSP camera, so it can provide more advanced video analysis functions for users. It can greatly improve the ability of video monitoring system, and make video resources play a greater role.
Panasonic makes the camera have intelligent function through the processing of DSP chip in the camera. Through the built-in intelligent processing algorithm, the camera can easily distinguish lost / dropped packets, detect moving objects and detect scene changes. After adding the intelligent function, the camera can easily distinguish the interference factors in the image from the actual moving target.
Sony recently launched a new generation of intelligent analysis function depa advanced. This technology completely moves the intelligent analysis and processing forward to the front-end network camera, the camera itself can analyze and judge the collected image, and transmit the alarm signal to the back-end system, which is different from the previous intelligent technology, which transmits the image to the back-end system and then analyzes the alarm by software, so the accuracy of the alarm is greatly improved. Depa second-generation intelligent analysis technology also has the latest research and development of “intelligent anti destruction detection” and “intelligent voice detection” functions. The intelligent anti damage detection function can recognize the actions of occlusion, focusing, spraying and displacement of the camera artificially, and send an alarm in time; Intelligent voice detection can alarm for squeal and abnormal sound.
Digitalization, networking and intellectualization are the inevitable trend of the development of video monitoring technology, and the emergence of DSP intelligent camera is the direct embodiment of this trend. In the process of the development of security monitoring system from digitization and networking to intelligence, in addition to the coding and compression technology using DSP chip, some intelligent function software of intelligence mostly need to be embedded in DSP chip for algorithm processing. Moreover, the DSP chip embedded in the intelligent software is embedded in the camera, which is called the detection and recognition intelligent camera. DSP chips are not only widely used in ordinary monitoring system, but also in intelligent video monitoring system.
Where is the development of DSP technology
In the face of the increasing social awareness of security, security products are constantly updated to meet the needs of different customers. Therefore, how will DSP, as the core device of camera, develop?
High integration has been an important problem, which can bring a lot of convenience and advantages to product design, production and use. But at present, the emergence of high-definition cameras and IP cameras brings a great challenge to the traditional DSP analog cameras. Due to the limitation of CCD IC pixels and analog video signal transmission, the improvement space of analog camera definition is limited. Ding Quan of Lejin Security Division said that in the face of market requirements, the special compensation function of DSP is very important. Strong light suppression, wide dynamic, digital slow shutter, digital noise reduction, electronic anti shake and other functions are particularly important. But at present, it is always a technical difficulty to deal with the time difference of the accumulator in the clock and kernel of DSP, so how to improve its function and effect is one of the processes of DSP IC design and development.
At present, the rapid development of digital network platform puts forward higher requirements for DSP. The traditional passive monitoring will gradually be replaced by the active monitoring with the functions of intelligent video analysis (video motion detection and tracking, human face recognition, human gait recognition, human and object abnormal behavior and illegal detention recognition, license plate recognition, etc.), According to the needs of different monitoring scenes, embedding the background video analysis into the DSP of the required camera can not only reduce the project workload and save costs, but also replace most of the work of people in the monitoring task. This is also the demand of intelligent monitoring direction for the development of DSP technology.