We have two parts in our project that we had previously submitted as part of the project.
2.Poster session information:
Poster presentation of the project is as follows:
This is our overall set up that it looks like:
When the customer enters into the trial room, these are the things that he sees on the mirror suggesting him/her to scan the barcode. Here is how the smart mirror (with its GUI) looks:
Upon clicking the Click here to scan option, the customer scans the barcode of the dress and the relevant information is retrieved from the database and it is displayed on the screen.
The customer can see the image of herself in different color of the dress. Image Processing concept has been introduced here where in the color that is available changes to the color that is required by the customer.
Here is another example showing image Processing where in yellow color changes to white.
QR Code is generated for this image.
If the customer wants to change the brightness of the trial room, then he/she could do the adjustment on the mirror.
The below video clearly explains the flow of the project from start to end.
The process demo is as follows:
The process barcode scanning and the light adjustment of the trial room is shown in the below demo:
The QRcode generation and the download of the image of the customer is shown in the below video:
Links to Software:
We’ve placed codes for the color change, QR code generator and the Lightning adjustment in the Github for which links are provided as follows:
Platforms for IOT projects:
Node Flow Editors:
Other useful links:
Project Sites used for my project:
The is a wide scope for improvements and additional features for the Smart mirror. We would have been implemented them if we had time, however if given a chance in the future, we would try to implement the below features.
1.The items the customer tried can be saved and from this data, data-analytics tools can be used to derive information about items that are most often paired with this particular item. This aids in giving appropriate suggestions to future customers and also in determining trends.
2. The mirror could be equipped with a quad-core processor, 2 GB of RAM, HD Camera, Wi-Fi module and Bluetooth, speakers, voice control and glass with anti-fogging and moisture. With the mobile application user will be able to select applications and widgets that will be reflected in the smart mirror and create profiles for each family members.
3. A cloud service could be used in the store so that the customer could request for an item from the trial room and the nearest sales associate would be notified and would bring the item to the customer.
4.A more personalized session can be created by having a database for customers and giving options for customers to save items for later.
It was really a great decision of mine selecting IoT from my electives as I had very good experience through out my journey. Firstly, selecting an IoT project was a big challenge for me. I have done research on many sites and came up with each project a day initially and was unsure of what to select from those. Being shopaholic, I was in an impression to do something that makes people comfortable without having any physical exertion and time saving to run across the store to fetch out the clothes that they had to try before making a purchase. Upon doing some research on this part in particular, we had come up with this Project Smart Mirrors.
Selection was fine but analyzing the requirements for the project was a big task and was not sure of the structure or flow of the project but we could figure out few things and kick started the project with the Bar code scanner part. I knew that Image Processing is a wider concept and overlapping the color onto the other is a challenging one. We had spent lot of time doing research on that and had to read a lot on other websites. After much hard work, the code was running successfully and I’ve added up QR code generator as suggested by my Professor. It sounded pretty simple and I was excited being able to download my images on the phone post scanning the QR code.
I felt setting up LCD was pretty simple but at the end, it was this part which turned up to have sleepless nights. We’ve installed all the required libraries and did exactly the same how it was suggested on other blogs but the display/touch wasn’t working. In this process of extending the Pi memory space, we got the Raspberry Pi accidentally formatted for which we didn’t have backup. So, we had to redo the whole process which was too tedious when there was not much time to do. Finally, was able to catch up the whole thing except that the images weren’t getting pulled up from AWS cloud for the QR code generator for which we had created a workaround to insert URL of the image into the code that we’ve written.
Finally, it was really a fantastic journey playing with electronic devices and I’ve learnt a lot from the project especially that there should always be a back up for the code. Images/videos of the demo that was working till date. If at all, we wanted to extend the space for the Pi, we should be having the iso backup for the OS so that it could be retrieved at any point of time.
Acknowledgements and References:
We are really thankful for being under the guidance of Prof.Paul Schragger throughout our journey. We referred “Practical Python and OpenCV” book by Adrian Rosebrock for the image processing module and ‘technoobsite.wordpress.com’ for the barcode scanning. We used TkDocs.com as reference to build the GUI.