product

Getting better and new day by day
"WooKyoung Information Technology" will provide high quality service.

Hi-VS (High Speed Video Searching)

It is a state-of-the-art AI-based video search solution.

System introduction.

A.I based High Speed CCTV Video Searching Solution

Hi-VS is a state-of-the-art AI-based video retrieval solution that derives accurate results with a high performance GPU based on deep learning based video analysis technology and algorithm.

Fast crime/accident prime initial action through high speed CCTV video searching.

View Image Larger

Object detection based A.I
  • Support searching height, cloth color, action and appearance of human object.
  • Support searching type, model, brand, color and number plate of vehicle object.
Increased processing speed using high-performance GPUs
  • Searching speed of 10~60 times speed
  • 1 hour video reference, maximum 12 minutes, minimum 4 minutes search time.
  • Support multi-searching for various video files
Available to compatible with the existed CCTV system
  • By compatibility with the existed CCTV, available efficient searching.
  • Decreasing video file download and copying time.

Features.

High Performance GPU
  • · Increasing processing speed using high performance GPU
  • · Support a video searching speed maximum 60 times
  • · Support multi-searching for various video files
Artificial Intelligence Technology
  • · Applying object searching and analysis technology based on deep-learning
  • · Provide high search accuracy for reliability
VMS compatibility Support
  • Available to search the specific video using VMS without file movement
Video File support
  • Available to search separate NVR and storage device by sharing files or storage with the searching server

Video analysis and searching feature based on A.I

Searching Speed
  • Detailed Search : 1hour video, required 12min
  • Standard Search : 1hour video, required 7.5min
  • High Speed Search : 1hour video, required 3.7min
  • Using Deep-learning Algorism based on High performance GPU
Object Type
  • Human, Vehicle
  • Build and learn self-learning datasets
  • Above 95% object accuracy (4096px above object)
Human Object Analysis : Gender
Vehicle Object Analysis : Brand, Model
Common Object Analysis : Color, Moving Direction
  • Cloth Color : more 8 colors / Vehicle Color : more 12 colors
  • Moving Direction : 8 direction, tracking technology

Hi-VS (High Speed Video Searching) Statistics.

View Image Larger

  • Moving Direction
  • Detecting Amount (Vehicle / Human)
  • Gender
  • Vehicle Brand
  • Color

DemoVideo