This article is about integrating OpenCV/PerC features with Microsoft Gadgets for making smart digital vision applications.
Applying For : Finance Category - Windows 8 Tablet
In highly populated cities, usage of self-serv / vending machines are usually high.
(Ex: On a salary day, In India, people would find most of the ATM’s queues filled )
People often make a queue and wait for the service. Examples include :
- Food & Beverage vending machine
- Ticket/Token vending machine
- DVD rental kiosks
On a busy day Vendors/Suppliers find it really difficult to know where to replenish first.
They operate solely based on the inventory level that is shown on their Remote Inventory tracker Application. This helps in a way but not totally smart.
The vendor would choose to go with C1 (Please refer the PDF) as the inventory level is critically low. As the existence systems suggest replenishment only based on the inventory level.
Though this is a helpful solution, it may not be the smartest one. Reasons below:
Smart Replenishing System
This uses a combination of OpenCV/Intel PerC SDKs with Microsoft Windows tablet tools & Azure to render an intelligent decision making system.
SRS uses digital vision to identify the number of people occupying the queue.
Camera could be mounted on a side/sides of the vending machine (or where ever suitable to track the peoples faces.
The camera reads the number of faces available in the queue and updates the information server on cloud with the vending machine ID.
The replenish command could be sent through Windows Tablet Consoles / Phones Controller App Manually.
SRS could also be configured to Auto-respond to replenishment needs. This service takes responsibility to send Replenish commands straight to Agents , bypassing the controller App.
While SRS is targeted LOB type applications for Banks, F&B vendors etc, Crowd Map is a subset App of SRS that helps a common user to locate the nearest "Less crowded" vending machine in his location.
It removes Agent & Controller App scenario from SRS and rather includes Direct to consumer App. It could be through Classic Windows App or even a Marketplace App. The client of this service (Here Bank is the client for the software vendor) determines the target software platform it wants to choose.
(Please check the attached PDF to see how this works)
SDK & Tools
Face detection module on Vending Machine :
C++ OpenCV /PerC for Face detection
C#.net Windows Services for Comunicating with SRS-OpenCV module to collect and report captured data to SRS server.
Server : Azure for Web Services , DB & Notification.
Controller & Agent App: Windows RT XAML Framework.
Keep a running update of any changes or improvements you've made here.