Notice
Back

Visual data analyzer that converts existing CCTV system 

into an artificial intelligence surveillance system in just 30 minutes

Deep Runner

Easy to build security services with a simple web interface

Unattended surveillance of up to 16 cameras simultaneously

Can also be used for factory quality inspection

If you have a brilliant idea to do with artificial intelligence,       how would you make it happen?

 Brilliant ideas like,
  • A portable device for blind. When a dangerous object or person is detected, the device informs the position and the size of the objects through the sense of the skin with electronic pulses.
  • An IT system that counts how many cows come in and out of cattle shed.
  • And millions other great ideas.
You may need to do the following:
It’s never easy.
  1. Hire deep AI developers, who may do the following things.
  2. Determine and purchase computer and GPU card to execute AI visual recognition algorithms.                       May be costly and can be a trial and error thing.
  3. Determine and install an AI platform such as Tensorflow.
  4. Select and test an appropriate deep learning algorithm.
  5. You may also need to train objects.
  6. Attach a camera to the computer and link it with the deep learning algorithm.
  7.  Develop a service scenario that monitors the recognition results in real time and processes when the target object appears.
  8. Develop user interfaces, case, and so forth . . .

The Deep Runner is a device that makes this complex task simple.

Using Deep Runner,
  1. Connect a camera to the Deep Runner and run it. Deep Runner recognizes 90 general objects by default. You can add more by training it.
  2. Insert the JavaScript code in your service webpage to access the Deep Runner, and run it using a web browser. The Deep Runner will be constantly watching the camera and will notify you immediately when the requested situation occurs.  

Basic function of Deep RunnerWhen you turn on the device, it recognizes objects from the camera input. You can check from HDMI video output.

Presence API : Notify immediately when an object appearsApplication examples: Notify immediately if suspicious person is found in security zone at night time zone. Notify immediately if a forest fire is found. (Note: recognition results appear before video due to the delay capturing Deep Runner output screen (HDMI).)

Presence API: Notify if an object stays for a specified timeApplication examples: If a person stays in the billboard for more than 10 seconds, the information of the person is notified. Notify when a car is parked for more than 10 minutes.

Absence API: Notify immediately when object disappearsApplication examples: Notify if the car is not in the parking lot at night time, Notify when treasure disappears

Absence API: Notify when object disappears for a given timeApplication examples: Notify if a guard is not seen for more than 30 minutes in guard post , Notify if an elderly person living alone is not seen for more than one day, Notify if livestock is not seen on the ranch for more than 10 minutes

Counting API: Notify the count of peopleApplications: Analysis of commercial power , Retail management (e.g. How many people at each table) , Parking management, Traffic Analysis

Counting API: Notify the count of people passingApplications: Counting people come in and out of a store , Counting cows come in and out of cattle shed , Traffic analysis

Multiple IP camera interworking :: Neurocoms Inc.Absence of bicycle is being monitored from the third camera in the first row and, presence of person is being monitored from the second camera in the third row. You can attach up to 30 monitors at a time.

Venture Company Certificate

Prime Minister's Prize at '2018 Korea invention patent competition'

Certificate of patent

Certificate of patent

Inquiry us


--
Submit