Inspiration

The human body pose points depict alot of information and this information can be used to solve and emulated many problems and even inev

What it does

The image(subject image) is sent on which orientation operations needed to be applied. The input images are stored on the server. Next the picture(controlling image) human pose with the hand gestures which actually emulates the commands for the change of the orientation the inputted image. This second image is sent to wrnch which returns the co-ordinates . Using these co-ordinates the backend logic generates the orientation command. The first input image(subject image) is processed as per the deducted orientation command.
The processed image is sent back on your mobile with changes.

How I built it

Image

Challenges I ran into

Initially wanted to impersonate the Human Pose by an Animated Character in real time. But ran into issue that python doesnt have any wrapper for unity animation engine and making it in any other language was not possible as per the short duration of this hackathon and being a single person working on a project.Then 3 members of my team left . So at last made a compromise of making it with the SDK to have some real time action with it but SDK wasnt working on UBUNTU 19.04 due to dependencies incompatible. So at last had to figure out by using HTTP Triggers.

Accomplishments that I'm proud of

Pulled the basic idea of my original idea even after running into so many barriers.

What I learned.

opencv image processing, trigonometric and co-ordinate geometry and its implementation in Python.

What's next for Lazy Hands Photo Resizer

Built With

Share this project:

Updates