Lazarus

4 mins read time |  1st Place in Challenge Android Machine Learning IBM Watson OCR Google Maps

The hackathon – HackUPC Fall 2017

October 2017 - Barcelona (Spain)

Over 700 students from over 20 countries will gather for 36h of dreaming, hacking and shipping on anything and everything. All projects are welcome, mobile, hardware, web, you name it. …

Hacking starts at 09:00 PM on Friday and ends at 09:00 AM on Sunday.

The solution

Lazarus, a mobile application for Android that helps blind or visually impaired people to move around the city and perform simple tasks of daily life without the face-to-face help of a caregiver, family member, or friend using only basic tools such as geolocation, voice recognition, or the camera.

Lazarus has two types of users:

• Blind users: people who need help in situations of confusion (like getting lost in a public space) or emergency (read a warning text, for example).

• Helper users: volunteers willing to “lend” their eyes to those who can not see. Its function is to answer the questions of blind people by transcribing texts or describing images

What can the app do for blind people?

Lazarus has five options that can be used manually or with simple voice instructions.

1. Location: A lost user can instantly know where he is thanks to simplified maps, specially designed for people with visual disabilities, and GPS.

2. Close to me: This option indicates relevant spaces that are around the user in an area of ​​500 meters as pharmacies, banks, post offices, supermarkets, or coffee shops.

3. Look for me: This option allows the user to ask other people what is happening around him, for instance, in situations of confusion or danger (knowing if a food is in a good condition or some pills without a box are the ones he has to take). The user only has to take a photograph and thanks to the help of volunteer users will solve his doubts.

4. Read text: Lazarus automatically translates written texts into voice messages through an OCR process.

5. Describe object: Lazarus can recognize objects from the real world in order to help you with your daily issues.

Languages

• Android
• HTML
• XML

Technologies

• Machine Learning
• IBM Watson
• Google Camera OCR
• Google Maps

Team

• Ester Lorente
• Sandra Martín
• Juan Salmerón
• Marc Vila

Hackathon result: 1st position in Zurich Challenge.

Role in the project: Android Developer.

Project code: Lazarus - GitHub.

Appears in:

• Winners HackUPC Fall 2017 - Zurich Challenge

• Els hackers de l’inLab son premiats a la HackUPC 2017 Fall Edition (CAT)

Hackathon website: HackUPC Fall 2017 - Devpost.