Sunny OIT 2019 CES VIP Session | Opening New Worlds with Machine Vision

On January 8, 2019, the 52nd International Consumer Electronics Show (CES) commenced in Las Vegas.

For this year's CES, Sunny OIT hosted an exclusive VIP session to provide clients with a more focused and enhanced communication opportunity. The session showcased a series of new products and supporting demonstrations developed around its depth vision modules.

Mars04 is an integrated module independently developed by Sunny OIT's Intelligent Optics division. It incorporates a Time of Flight (ToF) sensor, an RGB module, and an IMU module. Utilizing ToF technology to achieve high-precision depth information output, it enables high-frame-rate, real-time processing and output of 3D dynamic capture data, maintaining 3D point cloud output error within 1%. The module also supports automatic exposure and wide dynamic range (WDR), playing a crucial role in numerous vision application fields.

Dead angles, cleaning area limitations, and body size have long been pain points in mobile robot development. Sunny OIT's self-developed linear ToF module for floor-cleaning robots – Cleaner 01 – is the industry's first ToF 3D perception module applied in such robots for SLAM and obstacle avoidance. Compared to existing solutions, it offers significant advantages, effectively solving challenges like cleaning under low-height obstacles faced by the mobile robot market.

AI Technology Empowers Face Detection and Recognition

To meet the substantial demand for face recognition across multiple AI fields, Sunny OIT independently developed and exhibited several dedicated face recognition modules/complete units at this year's CES. These modules leverage AI technology for intelligent face detection and recognition. The latest ToF-based face recognition module supports embedded computation integrating three key modules: MTCNN for face detection, Facenet for feature extraction, and KNN for face recognition. Combined with Sunny OIT's proprietary embedded facial recognition algorithms, it provides users with an end-to-end intelligent face recognition solution, representing a high-performance edge computing offering.