DayBRIGHT: Image to Gradient Translation

The first translator I decided to fully flesh out for DayBRIGHT is the static image translator. The underlying system I have developed only has one principle rule - anything can be translated and used, as long as that translation returns a gradient object. So, how can a gradient be created from an image?

I took the route of writing an algorithm which finds and stores the most dominate colors in an image - then using each sample as a control point within a varying length linearly-smoothed gradient, where the most dominate colors are in order from left to right. Below are some examples of the results, where the top is the source that was sampled and gradient is the output visualized:


Fig 1.1: Sky converted to a gradient based on dominate colors


Fig 1.2: Composition VI, W. Kandinsky - image to gradient conversion


Fig 1.3: Composition VII, W. Kandinsky - image to gradient conversion

Now that a method existed to translate static images into gradients, it becomes a trivial task to write a translator for a video feed, as video can be easily treated as a series of images. Simple alterations to a subsystem can allow for a calculated sample of each frame in a gradient. So, if a video feed were parsed once every hour for twenty four hours, a gradient could be constructed of n-length, linearly being smoothed evenly across each respective control point based on n+1 distance.

DayBRIGHT: Generic Structures

The underlying system DayBRIGHT uses to translate various data structures has become less about constraining data to a set format, and more about developing a generic structure (a gradient) to fully represent the data in an elegant way. I find this an interesting topic to explore because it has really opened a lot of doors as to how far I can take his project; in particular, in the sheer variance of data I can now use.

RSS feeds, video streams, static images, audio tracks, physical sensor data - all of which can be harbored by a single generic data structure. Each piece of data simply needs its own translator towards being converted to the structure; once that is done - there is no further special-casing needed.

My goal for this project is to have, by the end of semester, a demonstration of how the various types of data listed above can be used to achieve results for an identical operation. In the case of DayBRIGHT, I will be using the gradient structures to light 3D geometry. How this will be done merits its own detailed entry - which is forthcoming.

I am really excited about this idea, it is a blend of engineering/mathematics and visual aesthetics that I have always been drawn to. I am dedicating all of my classes to it in hope that my time commitment will really push this idea as far as it can go.

DayBRIGHT: Proposal + Test Video

The idea behind the DayBRIGHT project is to be able to visualize data through HDR light probes being influenced by various types of datasets. Each dataset, be it an RSS weather forecast, webcam frame sequence, arbitrary color selection, or any other value - will be translated into a common format to be applied to a core HDR image.

The DayBRIGHT project will be encapsulated within a core software application which will communicate with various data feeds. As mentioned, each type of data will be converted to a common format: a color gradient. A color gradient will allow for analysis of RGBA, as well as time by sampling across its length. Depth can also be calculated by converting RGBA to luminance values.

To render the results, the initial solution will be to use 3Delight, an implementation of the Renderman standard, first introduced by Pixar. Using Renderman will allow for programming flexibility and high-quality renders at HD resolutions, if needed. An example of the early results is shown below, where an orange -> cyan color gradient has been applied:

DayBRIGHT: Gradient Sampling

One of the core concepts of the DayBRIGHT project that I am developing over this semester is to enable the ability to drive HDR lighting over time using a variety of methods - the first of which is using a simple color gradient.

For every frame to be rendered, a copy of the core HDR image is created, then a color overlay operation is applied to that copy - with the color applied being the current sample value of the gradient. This relationship is graphically represented below:


Fig 1.1: Graphical example of gradient sampling applied to image

Mind.Scribble.Form

My mind blurred, sketched, articulated, and scribbled on paper - most of which makes little to no sense to me currently. I normally never write things down to remember for the future - I generally only do so to calculate or brainstorm on the spot. Apparently, in parallel with my messy drawings, I also take photos without much care either. I think I'll write that off as part of my "creative process" and hope that you believe me (which you shouldn't).

Blast Radius

Several layers of particles revolving around a central sphere - each layer being rotated based on assigned attraction forces. Particles are symbolic of populous movement combined with the effect of unseen influence over collective entities - physically, socially, and otherwise.


Fig. 1.1: Straight-on shot


Fig 1.2: Corner angle


Source:
BlastRadiusMain.java
GLSL.java

Interactive and Time-Based HDR

I have been thinking a lot about pixels lately - most likely due to having a class with the given root word in it... twice. So, we have all of this information nicely packed within a frame - now what to do with it?

What I find interesting is how an image can be mined for data that is unseen upon first glance - and one of my favorite ideas in regard to this way of thinking is the creation and use of HDR images. High-Dynamic Range images are useful for an assortment of reasons - in the field of computer graphics, they are many times used as a light source. Paul Debevec is known as one of the groundbreaking researchers on the topic - and what initially began as technique for expensive high-end rendering has now been implemented in real-time in the past couple of years. Check out his site to see some of the amazing things HDR images can be used for.


Fig 1.1: Example of a light-probe HDR image
Credit: http://www.debevec.org


So, what else can been done with HDR - and how can it be used to create something interesting? For me, I want to explore moving away from leaving HDR lighting as a static image - and instead have it evolve over time while lighting a static scene. A few ideas are: time-lapse HDR lighting, real-time interaction with the HDR image itself, tying HDR colorization with weather forecasting, integration with existing data sets, and so forth.

Programming, Ubiquity, Devices, and You

After telling fellow students at ITP about the classes I decided to sign up for the coming Spring semester, I have often been asked the following two questions:

1. "Why are you taking programming centric classes if that is already one of your strengths?"

2. "Why aren't you taking any physical computer related classes?"

The answer to the first one is rather concise; I feel that staying constantly involved in software and mathematics allows me to build myself as a better overall problem-solver on both an analytical and creative basis. Writing code to solve a problem is concurrently a step-by-step breakdown of the problem itself; patterns emerge and reveal themselves, woven with systems of varying elegance and complexity. Understanding concepts which reveal such harmony will endure far past simply designing the next iPhone - leading me to the answer to the second question:

My position on physical computing at it pertains to projects at ITP (and in general) is that the development of devices which we physically interact with is inevitably obsolete. This is the century of biology - and consequently, biological micro-interfacing will be the future of device technology. Increased ubiquity leading to result/reward has nearly always been the trend in tool advancements - computing is to exist behind a curtain whenever possible. While it is certainly a valid venture to develop large-scale projects involving common electronics and so forth - I personally have little interest in pursuing similar implementation methods.

Software for Physical Display Project

Group: William McDonald, Sunghun Kim, Matt Parker

Further documentation:
Sunghun with videos / images
Matt on construction / mechanics

For our group project, we wanted to develop a physical system to represent graphical depth data. We decided to use a collection of sine and noise waves to start off with - as using depth data really only makes (practical) sense with a 3D grid.

Using Processing, a straight-forward GUI was created featuring four buttons, with a graphical representation of the waves being dynamically created to the right of the buttons. The fourth button allows the user to paint his/her own wave (or any shape, for that matter) and the software will generate a visualization and physical output for the image painted.



Fig 1.1: Drawing canvas allows user to draw images to physically visualize



Fig. 1.2: Visualization of user input wave drawing


To sample each wave - an algorithm was written which samples across the width of a given image proportional to the resolution of the physical system - in our case, 20 samples. At each sample - the program checks to see how far down (Y) the image it takes in pixels before a color other than white is found. Upon finding a color, the value of that sample is received by the simple equation of :

sample_value = y_coord / (image_height / max_value)




Fig 1.3: Graphical representation of sampling method on wave images


In regard to feeding this data into the servo motor array - we decided to use four arduino microprocessors connected via USB to the computer running the software application; with each arduino effectively controlling five servos:



Fig 1.4: Computer / Arduino / Servo array configuration

Shiver - Final

Fig 1.1: Splash screen for Shiver

The last week of work on Shiver has primarily been a focus on adding features on the GUI side and tying up some loose ends. The largest new feature is the ability to 'simulate' - which allows a user to simulate a full rotation (or however many wanted) of the globe, outputting the frames as an image sequence. This process allows the user to see the rotation as a movie clip running at full speed. The video in this blog entry is derived from that process.



Fig 1.2: Simulate options window


Shiver is still very much a work in progress - however, I am pleased with the result of my work over the past 6-7 weeks, particularly with knowing that I was able to achieve high-quality imagery in real-time using Python and OpenGL together.



shiver_simulation.mov - 10.6mb


Images from Shiver: