Dahua launches efficient IP camera series with built-in AI facial recognition

Jul 16, 2018, 11:12 am EDT | Stephen Mayhew

Video surveillance company Dahua Technology has launched a series of IP cameras with built in face detection and facial recognition powered by deep-learning artificial intelligence algorithms for real-time monitoring. The network cameras are focused on efficiency, taking pictures rather than continuous video for reduced bandwidth requirements, and the built-in AI automatically discards blurry or angled images.

The deepsense series cameras leverage Dahua’s wide dynamic range (WDR) technology, and provide high clarity facial image snapshots, even in heavily contrasted lighting conditions, according to the company. The AI also supports video extraction based on six different facial attributes, including gender, age, expression, glasses, face masks, and facial hair, to make searches and tracking more efficient.

The cameras also use advanced H.265 and H. 264 video encoding standards to save up to 70 percent in storage space compared to standard video compression, while maintaining high quality video. The built-in database can hold up to 10,000 face images, and deepsense IP cameras also support black/white lists and stranger mode, in which unrecognized faces can be treated as black-listed. Users can also adjust the similarity threshold based on their specific needs.

The company, which is based in Hangzhou, China, is marketing the cameras to business parks, schools, airports, banks, and shopping malls.

Startups Horizon Robotics and Kogniz separately announced the launch of cameras with embedded AI for facial recognition in April.