Please enable JavaScript.
Coggle requires JavaScript to display documents.
, ImageBay, - Coggle Diagram
-
ImageBay
Programming Tools
-
-
-
Bootstrap
Used to make the front end easier to design, and look more professional
-
-
SQL
Used for all functions relating to the database, or storing data
APIs
Google Vision
-
-
Detects labels based on AI image recognition, and therefore is not accurate 100% of the time
-
-
-
-
-
Usability Principles
-
-
Accessibility
How accessible is the solution, can it be operated on every device or browser?
Safety
How safe is the solution, particularly pertaining to the safety of the user's data?
Effectiveness
How effectively a user can use this website to get the desired outcome. How well does the website fulfil its purpose?
Restrictions
The cost of an API is dependent on the number of uses. Using the API may not be sustainable unless monetary generation is significant.
Additionally, hosting a server for this site to be online constantly would also incur cost.
-
-
-
Front End / UI
-
Must have an intuitive layout, which makes it obvious to users how the site operates
-
Forms should allow the front end to connect to a python back end, allowing for user interactivity
Must follow CARP design principles, and have a consistent design / colour scheme
Carp Principles
Contrast
Objects, images, text etc. All of these elements must possess contrast to stick out from the page and keep the users attention
Alignment
Border schemes must be consistent, and different elements of the website must line up. This provides a better user experience
Repetition
Repetition reinforces certain elements for the user, and gives them a sense of familiarity
-
Task
-
The application should allow users to upload images under their name, from which labels are identified using artificial intelligence. Images and their labels should then be stored in a database.
Users should then be able to search images by their detected labels through keywords. To accommodate for a wider user base (being a global company), ImageBay should allow users to search in multiple different languages.
Data
-
Database structure
-
Labels table
-
image id (fk, links to image_file table)
-
-
-
-
Image file table
-
email (fk, links to users table)
-
-
-
-
Success Criteria
Self Determined
4 pages, including: login/signup, home, upload, and search.(FR)
Must record user information through a login / signup function, then record who uploaded each image (FR)
-
Must take the user group into consideration, for example: translate the labels into French and German (NFR)
Must have a design which is aesthetically pleasing, but not overwhelming (NFR)
-
Must have an automatic language detection function, which allows users to search the database using keywords in any language (FR)
All of the applications function as intended, without error. (FR)
Prescribed
-
Must use Flask, Bootstrap, SQL, and Python (FR)
Must use the Google Vision API to detect labels in uploaded images, then store them for later searching (FR)
Use the Google Cloud Translation API to translate the recognised image contents into at least 2 different languages and allow users to search by those languages, particularly French and German. (FR)
Save the uploaded image, filename, and labels in a relational SQL database (FR)
-
When the user uploads an image, they should receive the detected labels before they are stored in the database (FR)
Responsive web design, including interfaces for: selecting a file for recognition, displaying of image recognition data for selected image file, and finding images for keyword search, then displaying the image (FR)
The web app complies with Web Design Standards, the Australian Privacy Act (1988), and the Australian Accessibility Standards (NFR)
Users
Personas
Jack Dickens
Graphic designer from Brisbane, Australia
-
At times needs images with certain content, but finds it hard to find the right images quickly as the uploaders sometimes do not use good or descriptive keywords
Michael Smidt
Graphic design student at the University of Hamburg in Germany often needs images for his creative work
His English is not very good and sometimes struggles to find the right images as he does not know the correct word in English
-
Fleur Boucher
Web designer from Paris, France
-
-
-
Global audience should be catered for, as the task is to create a website for an international company.
-
-