This paper first explores the technical characteristics of AVT (Advanced Video Transfer) technology in the context of coaxial high-definition digitalization. It then presents relevant system solutions and practical implementations, and finally outlines the future potential of AVT in enhancing the security of vehicle vision applications. Car video surveillance has become a critical application area within digital video surveillance technology. Due to factors such as vehicle safety and driving conditions, the video and vision systems used must be stable, reliable, and capable of handling strict transmission requirements. The digitalization of video transmission is now an inevitable trend. AVT, as a digital coaxial high-definition video transmission technology, offers superior image quality and robust system performance, making it ideal for supporting high-definition video equipment in vehicles.
The lossless and real-time transmission capabilities of AVT are particularly beneficial for extending video recording to advanced driver assistance systems (ADAS). This paper provides an analysis of AVT’s technical features in the context of coaxial high-definition digitalization, introduces related system solutions and their implementation through practical examples, and discusses its promising role in improving the security of vehicle vision applications.
The concept of car vision is becoming a reality. As video surveillance expanded into new areas, traffic became a key focus for industry growth. In addition to road-based surveillance, car video equipment is increasingly being installed inside vehicles to capture and process video data in real time, marking the beginning of vehicle monitoring. Early car video systems were primarily video recorders, designed to capture images around the vehicle during driving. These systems typically had a single front-facing camera with limited delay. To expand the surveillance range, rear cameras were added, increasing the number of video channels that could be processed separately or combined at the front-end processor. At the time, there were no strict requirements for signal transmission methods or quality.
With the development of intelligent video processing technologies, the video captured by in-vehicle cameras can now be analyzed in real time, providing early warnings or alerts for potential driving situations. This allows vehicles to have basic visual functions, enabling assisted driving. Car video systems can be broadly categorized into two types: 360° panoramic driving systems, also known as surround vision systems, and rearview monitoring systems. The components and video transmission methods involved in these systems are discussed below.
As shown in Figure 1, the 360° panoramic driving system consists of cameras located at the front and rear of the vehicle, each using a fisheye lens to capture wide-angle images. These images are sent to a host computer for processing, including splicing, trimming, and generating a 360° panoramic view, which is then displayed on an LCD screen. Currently, most 360° panoramic products on the market rely on standard definition solutions due to limitations in the main processor's capability. However, as more powerful processors become available—ranging from dual-core to multi-core—the performance gap is being closed, allowing support for HD video.
The resolution of car screens has also evolved, moving from 1280x600 to 720P HD. This shift has increased the demand for high-definition monitoring. Each camera must transmit high-definition video signals reliably via cables, making it essential to adopt high-capacity and high-reliability digital high-definition video transmission solutions.
Figure 1: 360° Panoramic Driving System
The rearview monitoring system represents a significant evolution in the design of driving recorders. Initially, these systems were tailored to meet the needs of different users with varying specifications and configurations. Before the rise of 360° panoramic systems, the demand for rearview monitoring was extremely high, with monthly market capacity exceeding one million units. The solution was cost-effective and easy to install, requiring only a small rear camera. When reversing, the display automatically switches to the rear camera image, helping drivers reverse safely. As high-definition displays become more common, more car owners are demanding high-definition cameras, creating new opportunities for digital HD in automotive rearview applications.
With the growing number of products in the market, more manufacturers are entering the competition. For example, Quanzhi focuses on T3 and V series chips, MediaTek launches intelligent rearview video chips, Mercedes uses the core 1860 chip, Rockchip and Intel jointly develop the Sofia3G-R platform, and other players use Qualcomm platforms.
Innovative approaches to video transmission in cars are reshaping the industry. Digital high-definition video transmission technology plays a crucial role in enhancing the performance and quality of car video equipment. It not only improves image clarity but also enables visually lossless and real-time video transmission. Both front and rear installations require efficient and reliable high-definition video transmission solutions.
Inside a vehicle, the wiring is complex and requires high reliability, leaving little room for flexibility in video connections. Ordinary AV cables are unsuitable for automotive environments, so coaxial or Ethernet cables are often considered. Coaxial high-definition transmission systems are currently popular, with analog solutions like AHD, HD-TVI, and HD-CVI widely used. However, digital coaxial HD systems, such as AVT, offer distinct advantages.
HD-TVI, HD-CVI, and AHD are analog high-definition solutions based on coaxial cables. They enhance image quality by increasing intra-frame frequency, resulting in improved resolution compared to standard definition. However, this method causes color interference and is less suitable for machine vision applications, where accurate video transmission is essential for ADAS and autonomous driving technologies.
AVT, as a new digital coaxial transmission technology, combines all-digital architecture with digital signal transmission, overcoming the distance limitations caused by signal attenuation. It uses a more efficient compression algorithm than standard VC-2 LD, achieving video delays as low as one frame period (under 3 milliseconds). This ensures visually lossless transmission, with no noticeable delay or loss to the human eye, effectively achieving real-time performance.
The visual lossless compression technique used by AVT achieves a signal-to-noise ratio of less than -40dB, comparable to uncompressed digital transmission. Image quality remains consistent regardless of distance, wire type, or temperature. In contrast, analog coaxial high-definition systems require conversion between digital and analog signals, leading to video loss and poor image quality. Additionally, analog signals suffer from nonlinear attenuation, further degrading image quality. Another issue is the large jitter in the sampling clock recovered at the receiving end, causing inaccurate sampling and inconsistent image quality across devices.
AVT also features rate-adaptive transmission technology, adjusting the compression ratio based on signal distance and attenuation to maintain optimal video quality. This makes high-definition video transmission independent of environmental conditions and supports future product upgrades. AVT also supports bidirectional data transmission, including UART, IR, I2C, and SPI interfaces, as well as SPDIF and I2S audio formats. It can transmit high-definition video over Cat 5 Ethernet cables, even supporting four video streams over a single cable, making it ideal for car wiring and the development of ADAS systems.
When integrated into a 360° panoramic parking system, AVT is illustrated in the block diagram of the system. Based on AVT technology, the NS2520 chip acts as the transmitter, while the NS2521 serves as the receiver. The NS2520 encodes parallel signals from the camera ISP, supporting multiple video formats such as BT656/1120, CEA-861, CPI, and DVP, with a maximum resolution of 1080P@30fps. The NS2521 can work with two NS2520s to fully support these formats.
Figure 2: Block Diagram of AVT in the Panoramic Parking System
The future of assisted driving looks promising. With the intelligent development of the automotive industry, ADAS will become widespread in cities, and various ADAS-related video systems will continue to emerge. These include adaptive cruise control, lane keeping, forward collision warning, lane departure warning, brake assist, rear monitoring, night vision, driver fatigue monitoring, and traffic signal recognition. These systems require high-definition, real-time video to ensure clear image quality and reliable transmission, which is critical for ADAS applications.
Traditional analog coaxial cables are prone to electromagnetic interference, making them unsuitable for high-definition video with resolutions up to 1080P. Digitization and networking allow video compression and transmission over Ethernet, but introduce network latency. AVT, as an advanced video transmission technology, uses digital coaxial high-definition transmission, offering low-loss and low-latency video transmission over Ethernet. This makes AVT an ideal solution for the automotive visual market.
As shown in Figure 3, AVT HD cameras are used in ADAS systems. The AVT chip is integrated into the HD camera, with the CMOS sensor and ISP processing unit highly coupled with the AVT transmitter chip NS2520. The video is seamlessly connected and transmitted directly over Ethernet. On the host side, the NS2521 receives the signal, restores it to a visually lossless digital HD video with minimal latency, and processes it in real time. The results are displayed on the screen and sent to the ADAS control unit for real-time adjustments.
Figure 3: Application of AVT HD Camera in ADAS System
In recent years, the automotive video market, dominated by smart rearview mirrors and smart driving recorders, has experienced rapid growth, with shipments reaching tens of millions. The convergence of in-vehicle video products is evolving, transitioning from SD to HD, from analog to digital, and from single-function to intelligent multi-function systems. Digital HD video transmission technology plays a crucial role in this transformation.
This year marks the first year of China’s smart driving assistance system market. ADAS will be primarily used for monitoring, early warning, braking, and steering tasks, with demand expected to grow rapidly over the next decade. It is a vital safety measure to protect drivers and reduce accidents, driven by both regulatory requirements and consumer interest. Internationally, the EU and the US have mandated that all vehicles be equipped with autonomous emergency braking and forward collision warning systems by 2020. Recent surveys show that car buyers are increasingly interested in ADAS systems that offer comfort and economy similar to those provided by parking assistance or blind spot monitoring.
With processors and sensors expected to dominate sales, semiconductor companies are competing by offering distinctive products and complete system solutions, not just for post-installation markets but also for pre-installed systems. The AVT series chips have undergone rigorous design and testing to meet vehicle regulations. AVT will launch 4K x 2K chips later this year, fully supporting ultra-HD ADAS requirements.
Infrared Optical,Zinc Selenide Lens,Optical Zinc Selenide Lens,Optical Imaging Lens
Danyang Horse Optical Co., Ltd , https://www.dyhorseoptical.com