The world's first network camera brand, AXIS from Sweden (market share of 69%, Frost & Sullivan, 2001), officially launched the world's first dual-mode network camera, AXIS 210, at the IFSEC exhibition in the UK in May, heralding the advent of the Motion JPEG and MPEG-4 dual-mode era. This also ended the long-standing debate over which compression technology was superior.
Based on the latest ETRAX system chip and ARTPEC-2 video compression chip from AXIS, the AXIS 210 network camera has the capability to provide services in both Motion JPEG and MPEG-4 (Part 2, Advanced Simple Profile at level 5) compression modes simultaneously. It can achieve processing performance of 30 frames per second at any resolution, up to 640x480 pixels.
In the past, Motion JPEG (hereinafter referred to as MJPEG) and MPEG-4 were the two most commonly discussed and applied compression technologies, each with its own strengths and weaknesses.
MJPEG is a dynamic image compression technology derived from the static JPEG image compression technology. MJPEG generates sequential moving images without considering changes between different frames within the video stream; each frame is compressed individually. This compression method is very suitable for static images, with resolutions ranging from 352x240, 704x480 to 1280x1024. Traditional JPEG compression processes an entire frame and requires waiting until the entire compressed file is transmitted before starting to decompress into an image, which can take tens of seconds or even minutes for high-resolution images. Basically, video playback is smooth if it achieves 30 frames per second.
The new MJPEG uses progressive transmission technology, first sending low-resolution images and then supplementing detailed data to improve the image quality. MJPEG compression technology can obtain very high-definition images, and the quality and resolution can be customized for any frame rate. The format can also capture individual frames, making editing easy. MJPEG's intra-frame compression method makes it suitable for video editing, as no decompression is needed, and only simple streaming is required, which consumes fewer hardware resources and reduces latency. MJPEG is an open format, meaning there are no licensing fees, and it has numerous developers and users.
Compared to other compression formats, MJPEG has low compression efficiency (see Figure 4). MJPEG's algorithm compresses each frame individually, rather than compressing based on differences between adjacent frames, resulting in significant redundancy and increased storage space—each frame can take up to 8-15 KB. High compression ratios can compromise video quality. Additionally, MJPEG does not support synchronized audio transmission.
Based on MPEG-1 and MPEG-2 compression technologies, MPEG-4 was developed as the latest MPEG standard for real-time transmission of audio and video signals over the internet, especially for mobile communication devices (e.g., third-generation mobile communications). MPEG-4 uses motion-compensated predictive coding to compress and transmit movement instructions and data, adjusting compression methods dynamically according to spatial and temporal features of objects, achieving higher compression efficiency compared to MJPEG and better image quality in limited and low bandwidth scenarios.
MPEG-4 can use existing transmission standards, including MPEG-2 Transport Stream, allowing MPEG-4 streams to be transmitted via existing MPEG-2 equipment, enhancing its market growth potential. It also supports IP network transmission. MPEG-4's compression differs significantly from previous generations. While MPEG-1 and MPEG-2 compression involved breaking images into blocks and handling repeated parts, MPEG-4 analyzes the foreground and background separately, compressing only dynamic elements, providing better compression efficiency and image quality. It has a flexible processing rate and offers a compression efficiency of up to 200x (800x for static images), supporting resolutions from 320x240 to 1280x1024 pixels. MPEG-4 is suitable for low-bandwidth, high-quality compression, interactive operations, and the fusion of natural and artificial objects. MPEG-4 also allows video units to be tailored to the decoder's capabilities, making it highly adaptive and scalable.
Due to MPEG-4's complex algorithm, there can be issues with interference and image distortion. Furthermore, as it uses inter-frame compression, MPEG-4 does not react immediately when objects are in motion, resulting in higher latency compared to MJPEG. The main drawback of MJPEG is its large storage requirements and bandwidth usage, but it remains mainstream in the DVR industry. As network and computer technologies develop, disk capacity and network bandwidth continue to grow, and users' focus will shift to image quality, which is not MPEG's advantage. Moreover, MPEG-compressed images are currently not accepted as reliable court evidence in Western markets.
There are many versions and levels of MPEG-4, which may not be compatible with each other, increasing development and integration costs. Licensing fees are also required for compression and decompression.
The dual-mode network camera combines the advantages of MJPEG and MPEG-4, allowing them to complement each other's weaknesses. In addition to the same unicast transmission mode as MJPEG, MPEG-4 also supports both unicast and multicast transmission modes. In a LAN environment, with multicast-supported routers and switches, using MPEG-4 for viewing does not increase network traffic as camera numbers or viewer counts rise, achieving real-time broadcasting in the LAN. If targeting internet viewers, using MJPEG with live broadcasting allows sending video to thousands of people simultaneously.
Given stable bandwidth, MPEG-4 viewing can significantly save transmission bandwidth, making it ideal for scenarios with limited bandwidth but requiring higher frames per second, such as traffic monitoring. On the other hand, with sufficient bandwidth, MJPEG can maintain high image quality, suitable for vaults, counters, or key entrances.
For recording, MPEG-4 can be used to save disk space, while MJPEG can be used to record more critical footage during alarms or motion detection.
Motion JPEG and MPEG-2 have been widely adopted and recommended standards in the industry, offering high-quality video. However, as MPEG-4's standardization and compression technology grow rapidly, replacing MPEG-2 as the market leader seems inevitable. AXIS has developed dual-mode products combining MJPEG recording with MPEG-4 viewing, which holds great potential for the future surveillance market, marking the advent of the dual-mode era.
蘊藏許多助人的知識與智慧。