Video Compression Reaching New Heights With H.265
Video surveillance equipment makers are focused on new technologies to improve image quality while conserving storage space. As a result, many are speeding up adoption to the successor of the reigning compression heavyweight champ, H.264.
TO MEET THE CHALLENGES of ever-increasing deployments of high-resolution video surveillance systems, High-Efficiency Video Coding (HEVC), or H.265, is quickly moving to the forefront of the industry. The greatest strength of H.265 technology is higher compression rates that reduce the design flow rate in order to lower the cost of storage and transmission.
H.265 implements the advanced techniques to increase bit-stream rates to promote coding quality and to create an enhanced relationship between time delay and algorithm complexity to achieve the best optimizations.
Video surveillance manufacturers are layering their own compression algorithms on top of industry standards like H.265. The goal is to optimize storage and bandwidth by recording higher-resolution video only where certain activities are detected. If nothing is happening within a scene except trees swaying, video data might be captured in low resolution. But if humans or cars are detected, those frames might be recorded in higher resolution.
Dahua Technology is an example of a vendor that is transitioning its video surveillance portfolio to ensure its products are future-proofed when the industry and end customers are ready to fully embrace this latest compression standard. As of November 2016, all of Dahua’s new product releases included dual codec support for H.265 and H.264 for backwards compatibility. Additionally, some products now include Smart H.265+, featuring a collection of intelligent encoding algorithms developed by Dahua as an optimized implementation of the H.265 standard, offering an additional bandwidth savings of 20-30% as compared to standard H.265 compression.
To ensure the standard can be used in products across different platforms, Dahua has built flexibility into Smart H.265+. Operators can configure the codec to tailor the encoding for a particular application or scene, adding flexibility and sophisticated image enhancement. Smart H.265+ technology is said to reduce bitrate and storage requirements by up to 80% compared to H.264 video compression, vastly improving the overall system capital expenditure (Capex) and operating expenditure (Opex).
Let’s take a deeper dive into Smart H.265+, which uses a scene-adaptive encoding strategy, dynamic group of pictures (GOP), dynamic region of interest (ROI) and intelligent noise reduction to deliver high-quality video without straining the network.
HIGHER EFFICIENCY, LOWER COSTS
Typical video surveillance applications require 24/7 recording, consuming a large amount of storage space. With the push to HD/UHD video resolutions with high bitrates, recording and storing surveillance video has become one of the main challenges facing the surveillance industry. The cost of transmitting large amounts of data while maintaining video quality and storing this data has exploded, increasing the cost of the entire surveillance system.
To resolve these issues, many manufacturers are continually investing in new technologies and innovations to reduce the cost of video surveillance systems. Vendors can achieve cost savings by upgrading their product lines to support the latest encoding standards (including MPEG4, H.264 and H.265) to take advantage of the improved transmission rates and reduced storage rates that these high-efficiency encoding standards offer.
Incorporating these standards, however, usually required a hardware update, and it took time for manufacturers to introduce the latest standard to the market. In addition, development and adoption of a more efficient standard could take a significant, undetermined amount of time. With more and more devices running the H.265/HVEC standard, only time will tell when the next standard will be available.
In order to stay ahead of the curve, Dahua elected to bring to market a suite of software enhancements to H.265 prior to it succeeding H.264 as the recognized industry-wide standard. To take advantage of these supplemental algorithms, customers need only perform a simple software upgrade on a device running H.265. Following are explanations for these key improvements.
Scene-adaptive encoding strategy – A typical surveillance scene is not fixed; the scene is constantly in flux even though these changes may be imperceptible to human vision. According to the Just Noticeable Difference (JND) model, human vision requires a certain amount of change to an object in order for a difference to be noticeable.
The way human vision perceives an object depends on several factors. For example, the human vision perception is more sensitive to changes to indoor scenes than to outdoor scenes. Humans perceive an indoor scene transmitted at a 4-megapixel (MP) bitrate in the same way as an outdoor scene transmitted at a 1MP-2MP bitrate.
Smart H.265+ technology adopts an encoding strategy that takes advantage of this difference in human perception to improve encoding efficiency. Scene-adaptive encoding implements a differentiated strategy based on smart scene analysis that includes the consideration of light, noise, motion conditions and enjoinments. Scene-adaptive encoding is designed to not only ensure the human vision perception of the surveillance scene, but also improve the encoding efficiency.
Dynamic ROI – Surveillance operators are more interested in moving objects in the surveillance scene and less interested in the background area. With this background in mind, Smart H.265+ implements the strategy of automatically separating the moving objects in the scene from the background according to the motion in the video scene through video analytics technology.
After separation, Smart H.265+ adopts varied compression levels according to different regions of interest. Typically in surveillance videos, moving objects appear only in a specified period and last only for a short time. Smart H.265+ may increase the compression level for a background area and reduce the compression level for moving objects, thereby enhancing compression efficiency.
Figure 1 (left) shows a red outline around the moving objects, marked as F. The rest of the scene is considered the background area and is marked as B. In most surveillance applications, an operator’s attention is concerned only about the moving object (F); thus, the image quality of the moving object must be maintained while the image quality of the background can be reduced.
Dynamic ROI uses a differentiated encoding strategy, as illustrated in Figure 2, to reduce the overall bitrate of the scene while maintaining the high resolution of the moving object.
Figure 2 (below) shows the same scene as depicted in Figure 1; however, the scene is divided into the background region and the motion region. Notice that the background region encompasses a larger portion of the overall scene, as compared to the moving object, thereby contributing more data to the overall bitrate. Dynamic ROI differentiates between the two regions, automatically reduces the bitrate of the transmitted video for the background region, and maintains the bitrate for the moving object. The transmitted bitrate is substantially reduced as compared to the initial bitrate.
Dynamic Group of Pictures (GOP) –
A video bit stream is comprised of a group of pictures: either an I-frame (intra coded picture) or a P-frame (predictive coded picture). An I-frame is a picture that is coded independently of all other pictures; a P-frame contains motion-compensated difference information relative to previously decoded pictures. The size of an I-frame is inherently larger by several factors of size than a P-frame, and comprises a much largeer portion of the bitrate of a scene than a P-frame.
The GOP is a group of successive pictures within a coded video stream. Each coded video stream consists of successive GOPs. Encountering a new GOP in a compressed video stream ensures the decoder that no previous frame will be needed to decode the next ones.
Typical surveillance video is mostly static for a majority of the time, with minor scene changes occurring infrequently. Smart H.265+ adopts the strategy of dynamic GOP and inserts an I-frame in the video stream only when a change interrupts the otherwise static scene.
The length of the GOP can be adapted dynamically according to the requirements of the surveillance application. The dynamic GOP structure reduces the number of I-frames, thus reducing the bitrate needed to produce the expected quality of video (see Figure 3).
In a typical surveillance system’s implementation of a fixed GOP, the interval between the two I-frames is fixed, generally at 2 seconds. So, a typical security camera transmits an I-frame every 2 seconds no matter the state of the scene. Keeping in mind that surveillance video is mostly static for a majority of the time and that motion (or a scene change) occurs for only short instances, enlarging the GOP could effectively reduce the amount of I-frames sent in a given time period.
Dynamic GOP inserts an I-frame in the group of pictures only when the camera detects motion to ensure the video quality according to the requirements of the surveillance scene.
Noise suppression – Noise artifacts typically show in surveillance video when the scene is uncertain and the lighting is poor. The noise not only affects the video quality, but increases the encoding bitrate. Controlling the noise level is imperative to implementing a better strategy for video encoding.
The traditional noise reduction strategy is to implement intense noise reduction techniques on the entire video scene, but this technique can delete some vital video details during the process. Striking a balance between overall noise reduction and preserving video details is a difficult task.
Smart H.265+ uses advanced video analytic technology to distinguish between the motion region and the background region, then applies a different noise reduction level to each region. Smart H.265+ is capable of reducing noise and preserving the expected video quality in the region of interest, especially in night surveillance scenes. In this way, it cannot only reduce the noise, but also maintain the efficacy the ROI video quality while reducing the overall bitrate.
Ultimately, systems integrators are compelled to stay abreast of the industry’s transition to H.265 in order to best serve their end customers’ security, organizational and budgetary requirements. Keeping pace with additional enhancements to H.264’s eventual successor will only further position integrators as their clients’ go-to subject matter expert.
Security Is Our Business, Too
For professionals who recommend, buy and install all types of electronic security equipment, a free subscription to Security Sales & Integration is like having a consultant on call. You’ll find an ideal balance of technology and business coverage, with installation tips and techniques for products and updates on how to add sales to your bottom line.
A free subscription to the #1 resource for the residential and commercial security industry will prove to be invaluable. Subscribe today!