AR Final Project

 AR Final Project

Idea: So... for this week I did something risky that I said I wouldn't do which is try out creating an AR application in Unity instead of a lens in Snapchat. I really liked the Image Tracking project that I did with mapping scenes from the anime onto the manga panels to create a fresh way to enjoy the story. (Previous Project Demo)

It's my dream to continue this project and turn it into a full-fledged application with image tracking for the whole manga volume. So I decided to get a head start and spend this final working on figuring out how to do that in Unity.

Problems:

Oh boy were there some problems. I spent more time troubleshooting than I did working on the actual content of the app. The first problem I encountered was getting a black screen on my phone instead of a camera feed. This was solved through changing some stuff in the build settings but it was a lot of trial and error. As soon as I fixed this problem, I rejoiced for all of two seconds before the app crashed. I now had a new problem which was again fixed with trial and error in the build settings. Finally after hours I could actually work on the project.

Process: 

At the moment, there is only one image tracking implemented. This was pretty easily done by following tutorials online. Although, it's far from my desired functioning, I suppose it's a good test run to make sure the AR and Unity is working.

Execution: 

Video 1

 

Video 2

 

Things to Implement:

  • Multiple Image Tracking & Tracking Images at Same Time
  • Only Tracking when Image is present (if app is left alone with target it'll just project onto the nearest flat surface)
  • Tracking Precision (kinda shaky atm)
  • Tracking to different sizes of images. Right now the specifications of the image tracking target have to be exact.
Far Future Implementations:
  • Home/Selection Menu
  • Sound???
  • Different Regional Versions
  • 3D models?
  •  

Comments

Post a Comment

Popular posts from this blog

Hypercinema Week #4

ArtLine: This creates line art from an image. I found this model interesting immediately. I think it has potential for a lot of reasons. It could be used to expedite the animation process by creating amazing line art from real photos. It'd be good way to incorporate real locations or people into animation easily. I also think it could be interesting as a way for non-artists to convert their pictures into a drawing. It could also be used as a learning tool for artists to better see what parts of an image to emphasize in line art and where to shade. For my concept, I would upload a series of images from a piece of media and create a manga style page with the results in illustrator or photoshop. It'd be cool to see how quickly and easily we could make things like that with this. Style2Paints: This automatically colorizes line art. I think this would be an interesting part 2 to the ArtLine model. I would put a set of photos through ArtLine to create line art and then put those gene...
Read more

Feeling Patterns Assignment #1

 P5.js code I wanted to explore patterns that aren't really repetitive in an obvious pattern way but that we still try to make sense of. Specifically, I remember someone mentioning raindrops on a window as one of the patterns they noticed and it stuck out to me because raindrops on a window isn't really a repetitive sound. The size of the drops, the amount, and the time between drops all vary every time. So why then does it stand out as a pattern? Probably because we try to make sense of the new pattern as the sound evolves and over time it all starts to sound the same. It is consistent in it's inconsistency.  For this project, I decided to play around with the sound of raindrops and specifically creating patterns and then breaking them with new patterns. Each segment is composed of a pattern. For example, short length clips spaced out evenly, but then layered with more short length clips spaced out evenly but at a different start point and some long length clips spaced out...
Read more

ICM Week #7

For my midterm project, I wanted to make a rhythm game. I was uncertain about how exactly would be the best way to go about that, but I settled on a 3x3 grid with "notes" coming in from off-screen and the proper keys would need to be pressed when the note passed in front of the corresponding box. This was a challenging and long project. I'm proud of the coding even if I think from a game design point of view it could be better. The only thing that is "incomplete" about it is that I didn't feel like coding in all of the music cues so there's enough of the game set up to play about 40 seconds right now. Right now it's coded to use the keypad, which I'm not sure if mac laptops have, but it was more intuitive than using the letter keys. There is a volume slider, but if you change the volume it also messes with the beat detection that changes the background color. So if you turn the volume all the way down, the background stops changing colors. I don...
Read more