Virtual Shopping System With Supporting Gadgets
'Virtual Shopping System on a Personal Server with Supporting Gadgets' aims to introduce an innovative concept of Virtual Shopping. Some additional gadgets provide the best complement to make the system autonomous. System design proposes to bring all the actual marketplaces
2025-06-28 16:36:38 - Adil Khan
Virtual Shopping System With Supporting Gadgets
Project Area of Specialization Internet of ThingsProject Summary'Virtual Shopping System on a Personal Server with Supporting Gadgets' aims to introduce an innovative concept of Virtual Shopping. Some additional gadgets provide the best complement to make the system autonomous. System design proposes to bring all the actual marketplaces on land on our android application. By doing so, a customer does not need to go and search for a product shop-to-shop, instead, he can find and buy a product of his needs virtually by a cellphone application.
There will be different gadgets like a barcode reader and a camera circuit that will update the account of a shop keeper in real-time.
- Two android applications - one for the seller and one for the buyer will be developed
- Barcode reader will be interfaced with the system with our own-made software
- Product detection will be done in shops after finding an efficient object detection approach
1. FINDING THE BEST OBJECT DETECTION APPROACH
A part of our system is to detect different products in a shop and put them in an online database. Android app integrated with this database will then display all the detected products in a shop to customers. For this goal to be realized, we had to build a perfect image detector to update our database in real-time. Our team did research and tried different approaches of object detection which includes object detection with OpenCV library, google images back search and detection with Google's TensorFlow framework. Among all these different approaches object detection with TensorFlow showed the best results. Hence, we have done object detection with TensorFlow which is a Framework for deep learning (a subset of machine learning).
The next challenge was to select a processor which can run TensorFlow efficiently and is cost effective too. So, Raspberry Pi is chosen for this cause. After the training of different products was completed, the system was gone through the testing phase. The results achieved were of higher accuracy.
We tested the following approaches to find the best object detection method:
- D-Classifier (Object recognition using Reverse Image Search)
- Object Detection Using a Trained Neural Network
- TensorFlow (Selected Approach)
2. INTERFACING BARCODE READER WITH RASPBERRY PI
Barcode reader has been interfaced with Raspberry Pi using Python programming language. The python program is written completely from scratch. The main module used is pygame. The barcode scanner decreases the quantity of a product from the main database after scanning a barcode from it.
3. ANDROID APPLICATIONS
Two android applications - one for the buyer and one for the seller were needed to be developed. All the data of the android application is saved on MySql database on our personal server. Android application is connected with the server using the Retrofit library for Java. The transfer of data from the app to a server or from server to app takes place with the help of an API written in Php server-side language.
Benefits of the ProjectSome e-commerce giants like Daraz.pk, Shophive.com, etc. are working successfully in Pakistan but the products available on their websites are very much expensive as compared to those that are bought offline by market. The minimum delivery time of these services is no less than one or two days.
So, even after spending so many years they fail to provide a fast delivery service like Amazon in Pakistan that is a major drawback. Also, online selling is still considered unsafe, unreliable and not profitable in our country. On the other hand, people prefer to go to market themselves to buy their desired products. But shopping wastes a lot of time of Pakistanis which they can spend to do something productive. So, there is a dire need of a system that brings all the market-places on a smartphone app. So that a customer can order any product from that virtual market app and he gets the product in the next 2-3 hours without actually going to the market.
Since a lot of problems are still present in the e-commerce industry of our country. So, our team took this challenge and is trying to solve it in a given span of one year. We believe that our system will open new opportunities for wholesalers, traders, and retailers to sell their products online, providing them a new and more simplified way. It will give them relaxation from high commissions and delivery charges, the frustration of shipping and provide them social marketing facility to target the right customers for their products so that they could thrive! And our country could thrive!
- Different approached are tested for object detection and object detection using Google TensorFlow Framework is found as the most efficient amongst all. The code is written from scratch without copying it from anywhere.
- Barcode interfaced with the raspberry PI and integrated with the database. Software is made for this purpose using Python language.
- Buyers and sellers android applications are programmed on Android studio using Java language from scratch.
- A database designed for buyers and sellers application using SQL language.
- An API is written in PHP server-side language and all the requests are tested using postman software.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 30000 | |||
| Raspberry Pi 3 B+ | Equipment | 2 | 6000 | 12000 |
| 32 GB SD Card Class 10 | Equipment | 2 | 1300 | 2600 |
| HDMI to VGA converter | Equipment | 1 | 400 | 400 |
| Raspberry Pi Official Camera Module V2.1 | Equipment | 1 | 4000 | 4000 |
| Barcode Reader | Equipment | 1 | 4000 | 4000 |
| Printing progress report and final report | Miscellaneous | 1 | 1000 | 1000 |
| Shelf for products | Miscellaneous | 1 | 2000 | 2000 |
| Wires, screws, pcb, soldering, electronic components | Equipment | 1 | 2000 | 2000 |
| Products to display | Miscellaneous | 1 | 2000 | 2000 |