Hikvision has unveiled an innovative solution in video compression technology with its new Guanlan Encoding. This AI-driven advancement offers significant storage savings of 30% to 50% on average, all while maintaining the quality of essential video footage.
As the demand for higher resolutions, increased channel numbers, and extended retention times escalates, storage expenses have become a major financial factor in video security systems. The Guanlan Encoding system, based on the H.265 international standard and utilising Hikvision’s Guanlan Large-Scale AI Model, expands the application of this AI from video analytics into the core encoding process.
In larger-scale security settings, Guanlan Encoding can dramatically decrease the need for HDD capacity and required rack space, along with cutting down long-term energy use. This efficiency contributes to a reduced total cost of ownership throughout a project’s duration.
Traditional codecs treat all pixels uniformly, placing operators in a difficult position of choosing between high-definition image quality and efficient storage use. However, Guanlan Encoding resolves this by focusing on key objects within a scene, such as people and vehicles, preserving them with high precision through Region of Interest (ROI) segmentation, while compressing less important background data.
Jason Yang, Vice President of Hikvision International Business Centre, stated, “Storage efficiency has almost always come at the cost of quality – until now. Guanlan Encoding looks at the picture first – then decides what’s worth keeping in full clarity.”
This technology uses two synergistic modes: Dynamic Sensing, which dynamically adjusts the bitrate to sustain detail in rapidly changing scenes, and Static Optimization, which applies high levels of compression to still or low-motion sequences, reducing frames to minimal byte sizes when possible. This approach moves video encoding from a generic solution to an intelligent, demand-based model.
Hikvision’s internal tests demonstrate the effectiveness of Guanlan Encoding in saving bitrates compared to standard H.265 encoding. Savings reached 49% for 24-hour operation in a canteen and 42% during peak traffic at an office park entrance.
In a corporate lobby, savings were 38% over two hours, and in a bustling commercial street, 18% savings were achieved in just one hour. Fully compatible with current H.265 decoders and both Hikvision and third-party devices, Guanlan Encoding supports advanced AI analytics without altering encoding format, frame rate, or resolution, facilitating smooth integration into both new and existing systems.
This technology is incorporated in Hikvision’s DeepinView(X)-Series Network and PTZ Cameras, Ultra-Series Cameras, and Cameras featuring ColorVu 3.0, as well as DVRs.
Plans for further expansion to other product lines are underway, ensuring applicability across diverse environments including enterprise campuses, retail chains, public venues, and critical infrastructure. By embedding AI into the core of the codec, Guanlan Encoding represents a shift towards targeted video compression that prioritises meaningful content over mere pixel data.
