Translating

"Translating" is a collaborative project in which I am currently working with my friend David H. del Valle, From Universidad Pontificia Bolivariana Medellin. He is working on his College Thesis, studying  human flows in habitable spaces.

This project is still in a really early development stage, but i find it quite interesting, and since i had not  uploaded anything lately i decided to share with you this work in progress.

So, the part in which we are working now is to make this kind of "translation" from JPG images, PDF, and similar formats to Grasshopper and Rhino elements.

The idea here is to explore some ways in which we could translate real time motion capture to  our computational models so that we can study human flows in architectural spaces and get a visual feedback to create better architectural planning schemes.

The definition is really simple, it works based on the ImageSampler component. the problem here is that, as we are working with non-vectorial images, the numeric output is a really wide numeric range, and the point cloud is really a mess.

In order to solve this problem I developed a "cleaning node", which works purging the data,  analyzing the Z coordinate of each point in the cloud and taking just those on top of the image curves.

I still have some work to do in order to get a cleaner point map, given than the definition still has some problems dealing with straight segments of the image curves it is needed to have an enormous point density to get all the curve segments covered.

This definition generates two outputs, the first one creates the point cloud, and the second one just takes the numeric values from the grey scale of a picture and resamples it

so...here is Zaha !!!