/H.265 technology has extended encoder framework to enhanc...
It is understood that the foreign accounts leading H.264 video encoding technology to use more technology. The new generation video encoder, the same bandwidth, double the picture quality. The same quality video with H.264 technology blurry picture card phenomenon, “mosaic” accumulation, while opening the “H.265 starting experience” to watch the video clear and smooth. December 2013, China’s first real-time H.265 video encoder in Xingtai come out, which makes our country in the field for the first time as an international video standard products of the leader, thus ending the holding device can only watch high-definition standard definition history.
As we all know, video encoders are used to implement front-end signal (video, audio, and other signal) digital compression and network, specific features include analog video and audio information and access, encoding/compression, transmission of alarm information and peripheral devices (such as cameras, Yun Jing, matrix, etc) control.
Biggest difference is with network cameras, video encoders video source from the analog camera, need to be used in conjunction with analog cameras, network cameras are integrated, integrates analog video-capture feature in itself. From currently of situation view, despite network camera is large emerged, but because following two a reasons, video coding device still will in network video monitoring system in the occupy not alternative of important location: a is large has built of simulation and digital monitoring system urgently network transformation, to protection existing simulation camera of investment, these transformation will produced huge of video coding device deployment needs; II is currently network camera of select surface also far no simulation camera big, to meet different user differences of of application needs, So many applications must be based on the model of analog cameras and video encoders to achieve front-end digitalization.
Encoder for storage and video management become more efficient
According to the IMSResearch study found that at the time of steady development of simulation technology, IP video is 30% growth in stride a year chasing simulation market. Two different monitoring systems cannot be joint development? In addition to technical applications are different, in most cases, there is not much overlap between them.
The video encoder via analog camera input signal into a digital video stream, a common compression standards, such as M-JPEG, MPEG-4 or H.264 output video format will be the same, can be fully integrated into network cameras and network video systems. Codec is also built-in features also include video motion detection, tampering alarm and video analysis.
Monitoring in the market there are many kinds of encoders, single channel or multiple channels to connect analog cameras, should be installed independent of the encoder, close to the analog camera signal, otherwise it will affect the digital conversion result.
Video encoder make the network more flexible and more flexible.The expansion of the network is endowed with more options, in a new simulation of IP system, adding new camera will become simpler, thus simplifying the work of the system management services.Traditional analog closed-circuit television monitoring system of proprietary encoder will let you walk into a dead end.Network to establish open and interoperable standards provides the possibility of more.Maybe you don’t understand the technical principles, but will eventually feel the advantage of video encoder to save bandwidth and storage.
Through the encoder to simulate system into IP system, the biggest benefit is to save the cost and make up the technology gap.A new digital network applications, can be in any place, any time to realize the remote access monitoring images, improve image quality, storage and video management become more efficient.
H. 265 technology continue coding framework of enhancements
From the coding framework, h. 265 is still used for h. 264 hybrid coding framework, but ascension or improvement of each technical details.Larger improvements are: 1) in the image block and motion compensation, transformation, support the larger sizes and species;2) more frame/interframe prediction and motion vector prediction and transformation model;3) increasing SAO ring sampling adaptive filter;4) provide the TILE pattern, to better support parallel processing.Application of these new technologies, not only effectively enhance the compression performance, as well as various effectively expands the space of the processor platform.
1, more macro block, and transformation.Relative to h. 264 of 4 * 4, 8 * 8, 16 * 16 macroblock type, h. 265 introduced 32 * 32, 64 * 64 or 128 x 128 macro block, the purpose is to reduce the number of macro blocks hd video, reduce the parameters used to describe the macro block content of the message, plastic transform block size also expanded accordingly at the same time, to reduce h. 264 asked the similarity coefficient of transformation in the neighbouring blocks.
2, using the new MV (motion vector) forecasting method.Different from h. 264 motion vector prediction method based on space domain, h. 265 frame prediction, expand more direction at the same time will predict block set from spatial domain to time domain and space-time domain, pass rate distortion criterion to choose the best prediction block after calculation.Test, using this method, the fundamental mode in h. 264, under the condition of same quality with the 6.1% average compression gain, complex image compression gain can even increase to 20%.
3, more consideration parallel design.Current chip architecture has been gradually from a single-core performance multi-core parallel direction development, h. 265 introduced Entropyslice, WPP, such as parallel computing ideas, to a higher degree of using parallel encoding algorithm, is more advantageous to h. 265 in the GPU/DSP/ASIC/FPGA parallel degree very high CPU fast and efficient in the realization of industrialization.
4, the newly added Tile dividing mechanism makes the previous slice, frame or GOP as the unit of coarse-grained data parallel mechanism is more suitable for parallel implementation of isomorphism multi-core processors.Dependentslice and WPP mechanism has solved the previous h. 264 coding technologies such as the entropy coding link to parallel implementation problems, make the decoding process of DCT, motion estimation and motion compensation, entropy coding task module division more balanced, significantly improve the parallel speedup.
5, lower stream.Repeated quality comparison test has shown that, under the same image quality compared to h. 264, by h. 265 coded video stream size than h. 264 around 39-44%.Due to the different measurement method for quality control, the data will have corresponding change.By subjective visual tests data showed that the rate cut 51-74% of cases, h. 265 coded video quality also with h. 264 video coding approximate even better, its essence is better than expected signal-to-noise ratio (PSNR).
The encoder will become smarter
Encoder is one of the mechanical equipment components, with the improvement of automation degree, realize the signal conversion is no longer the only needs, users a higher level of integration, more durable products, and more abundant interface mode, etc., all of these will be the encoder to the intelligent process.In the domestic market, low-end products will be gradually marginalized, with their own consciousness and the market demand change, will be more high-end encoder, accepted by the public and use.What would be the market pattern of the future?We have to wait and see.