
Mobile PoS app with visual recognition
Mobile app for leading DIY chain stores integrated with the retail Point of Sales able to recognize the Items via computer vision and insert them inside the customer’s shopping cart.
The Challenge
The need, for the customer, to be able to uniquely and immediately identify products through a mobile application integrated with an image recognition system. The system would then have to show alternatives to the user and share the information securely with the point of sale checkouts.
The Solution
MOLO17 has developed a solution that consists of a Mobile App for Zebra TC8300 handheld devices and an integration layer for the proprietary LINUX-based cash system.
The devices frame and photograph articles through the devices, allowing the image recognition system to identify and find the article in question or suggest variants for users to choose from. From the mobile workstation, this information then goes directly to the till and this allows to speed up the information avoiding errors and speeding up the time.

Mobile App for Device
1.
Association of the cashier in the process of accessing the application. Subsequently, the user can change the till to which the device is currently associated at any time by simply scanning the QR code associated with another till.
Mobile App for Device
2.
Acquisition of the picture of an article: After associating the till, the user can start acquiring the pictures of each article. The photos to acquire have a predetermined height to width ratio to help the user to acquire the photo in the correct format avoiding subsequent editing.
Mobile App for Device
3.
Scan barcode of an item: the user can also scan a barcode using the physical button of the device in use.
Mobile App for Device
4.
Article selection: Once the system has finished processing the captured photos, the user displays the detailed card of the article that is closest to the photographed article.
In the detail screen they are present:
- Article name
- Article image
- Description
- Article code
- Specific variants e.g. color, number, etc.
- Button to confirm the article
- “Item list” button to return to the item list.
Mobile App for Device
5.
Article confirmation: The user, after having verified the correctness of the chosen article and the possible choice of the correct attribute, can confirm his choice. The system will then send the article code to the checkout while being able to go back in the operations in case of errors or changes.
Cash desk software
The checkout software is a Java 6.+ application consisting of two structural layers that allow integration with the normal software processes in place on the point of sale checkout management system.
Results
- drastic reduction in the time needed to identify and verify articles
- precise data acquisition for the receipt
- secure transmission of information to checkouts
- ease of use and training for existing and future personnel
We only know a fifth of what lives in the depths.Everything else is left to explore.
coderspace is coming.