Dahua Technology, a world-leading video-centric smart IoT solution and service provider, is bringing new AI features to its WizSense family. Adopting an independent AI chip and deep-learning algorithm, the newly introduced WizSense integrates SMD 4.0, Quick Pick, AI Scene Self-adaption (SSA), AI Coding and other extended AI functions to deliver accurate detection, fast target search, simplified operation and inclusive AI products to customers.
“User demands are becoming more and more diverse, with higher requirements for AI accuracy and better image quality in different scenarios,” said Eaden Xie, IP Camera Product Director at Dahua Technology. “That's why we have brought four major upgrades, including optimization of alarm accuracy, search efficiency, etc. We believe that the development of high-end WizSense will bring more value to our customers.”
SMD 4.0: Focus on real alerts and color information
Thanks to an optimized AI algorithm, SMD 4.0 futher reduces false alarms caused by large and small animals. Compared with SMD 3.0, the false alarm rate for large animals (cattle, horses, deers, etc.) is greatly improved from the previous 5% to 2%; while the false alarm rate for small animals (cats, dogs, etc.) is further enhanced from 1% to 0.5%. Filtering out false alarms caused by non-target objects enables users to receive valid alerts for human/vehicle targets that may pose a real threat and to respond as early as possible. At the same time, the powerful AI capability of SMD 4.0 can detect the color information of clothing and vehicle, like exterior color of a vehicle, top clothing color of a person, etc., which provides data support for evidence-based backtracking after an event.
Quick Pick: Quick search for human/vehicle targets
This year’s WizSense brings a brand new Quick Pick feature that comprehensively enhances the efficiency of post-event retrieval by 90% compared to SMD Quick Target Search. Quick Pick (with AI NVR) allows users to quickly search for the target of interest (human or vehicle) using AI algorithm-based extraction and comparison of clothing and vehicle color, rather than searching from massively categorized "human" and "vehicles" playbacks. It significantly narrows down the search results to only present the video footage of the targets selected by users, which greatly saves search time and labor costs. It is ideal for various verticals such as retail stores and shopping malls to quickly locate a specific target during or after an event.
AI Scene Self-adaption (SSA): Smart recognition of multiple scenes with optimized image quality
With its capability to recognize monitored scenes with changing weather and lighting conditions such as rain and fog, backlight, etc., the AI SSA technology of the new Dahua WizSense can intelligently adjust the camera settings (e.g. brightness, color, contrast and white balance) to capture clear and high-quality image output that matches the actual indoor and outdoor scenes. This way, installers no longer need to manually adjust various parameters of each camera – which often requires additional time and labor costs – to improve video performance. AI SSA supports multi-scene recognition, making it suitable for both indoor scenarios (e.g. warehouses, offices, museums) and outdoor sites (e.g. villas, schools, factories, shopping malls).
AI Coding (IP Solution): Save storage while ensuring the image quality of targets
The increase in monitoring resolution has prompted the demand for more storage space. Based on CBR (Constant Bit Rate), the AI Coding function of WizSense for IP Solutions dynamically allocates the bit rate to provide clear target images and saves bandwidth up to 25% compared with H.265. At the same low bit rate, WizSense AI Coding prioritizes bit rate allocation over people and vehicles to ensure clear and detailed images of human/vehicle targets, while saving storage costs for users.
With the addition of the aforementioned new features of Dahua WizSense, users can now receive instant alerts during an event, quickly review and search through volumes of SMD video footages for specific targets after an event, prioritize matching for the best image quality, as well as saving storage with AI coding, facilitating post-event evidence collection while adhering to privacy protection. Additionaly, it creates an opportunity for systems integrators to provide comprehensive security solutions through Dahua’s WizSense portfolio that can transform video into actionable intelligence while saving users’ time and money.