DONE and done!

Hah! When you least expect it, it is done! Did my presentation in Umeå today, all seems fine with my supervisors.

If anyone is reading this, do not anymore :)

Report almost done?

I never thought the report would take so long to finish. I've worked for three weeks with it now, and I'm nowhere near a finsihed report. I've had great help from my tutor at the university, providing me with fast feedback.

Writing report when the weather outside is great is very hard...

End of implementation, start of report

Ok, so making a warp filter was not as easy as i thought, thus has two weeks of implementation gone in a jiffy without any results to show. I'm not satisfied with that, but I hope I can finish it the coming week. After that I'll start writing on the report. People has told me it's gonna take ages getting it finished but I hope I'll get it ready in time to do my presentation 1:st of June. I'll start working on SEMC the 4:th of June, some pressure building up ye? :)

Getting closer to the real thing

Back from my week off, good times. Big difference between Abisko with over a meter of snow and Lund with almost summer weather.

Begun implementing face warping filters this week to make happy/sad/whatever faces on people. Quite cool. Also started looking at the report and presentation and all the things around that.

week_off("Abisko");

Just now I'm traveling by train to Abisko for a weekend of snow and heliski. Gonna be nice with some time off.

When I come back I'm gonna plan for new filters, make a midterm presentation for the department and attend an interview. Busy week :)

Some output ...

As promised I've now put some output from my work here. My fantastic super mega sketch-filter is applied on these images, first one is the original, second one with color blending and the last is the black and white version (no blending with original).






Not that fantastic, needs quite high resolution on input otherwise it will be very blurry because of the double 3x3 kernel filter pass. Still it's not that slow on a big image.

Next week will maybe bring a halftone filter and then an other attempt two strip Technicolor filter.

Weeks keep running...

So, week number seven has started and I thought it was time for an update. The weeks keep running and I feel that I've produced at least something, but not as much as I first thought I would.

I've been looking at HDR and Tone Mapping filters the last weeks, reading some papers and trying to implement a nice TM (Tone Mapper). Almost all papers included algorithms with nice result but none of them would be possible to implement on an cellphone because of the limited CPU. So, the best TM I implemented was the most trivial one... (pixelA + pixelB) / 2 :)
Ok, now I fine tuned the algo somewhat by adding support for non linear clipping of white values, but still very trivial. But I know a lot more of HDR and TM now.

The coming weeks are gonna be about sketch/line art filters.

Stay tuned!

Moving on

A few weeks has past since my last post. The weeks has passed with incredible speed. In the coming week we will try to move my engine to a phone, having worked in an desktop environment. Will be interesting to benchmark on a much slower CPU.

So far I've implemented the ability to run pixel wise filters and support for kernel filters of different size. Some different effects combining the different filter types in a chain. The main target has been just implementing and not optimizing, which will be necessary before we move the code to a phone.

I've had my doubts on my coding- and problem solving ability but summarizing shows that I've still have produced a working filter engine with a few filters and learned a lot about the subject of image processing by reading different articles and forum posts.

It's fun and challenging!

One week down, a million to go?

The first week has passed at almost no time at all. The week consisted of trying to figure out how the C programing language worked, not having used it for almost a year.
Have made some progress with the design of the filter engine and I begin to understand my task. Will move to a cubical from my fantastic office on Wednesday :( But those are the rules as a freshmen...

MT started

I've begun my MT work at SEMC this Monday. Started coding from the beginning, against all methods for a good work. But no worry, just implementing a simple test filter engine at first to later on implement really cool filters. But so far so good, feel very happy for this opportunity!

Moving to Lund

Moving on Wednesday to Lund. Got a room with two guys I've never met but they seem ok. I'm going to have my first meeting with my tutor Andreas on Monday the 22:nd at SEMC. Will probably set up some kind of time schedule for the spring and start reading books for a month or so.