I am trying to convert a 3d point from one space to another. I have a kinect cloud and i am drawing this after a few translates and rotates withing a popMatrix()/pushMatrix(), i do this so that the pointcloud is in the right space if the original origin is my point i want to draw and stl mode(but thats not important). Anyway.. after pusing my matrix and translating and rotating i find the centre of gravity of a person in the kinect cloud, but those x,y,z are relative to the new rotated/translated origin. I was under the impresion that calling model(x)/(y)/(z) would convert that point to the original co-ordinate system??? but when i do it it doesnt give me the right point. I then noticed that when i moved the camera with peasycam the converted point changes...
Maybe i dont understand the Push/popmatrix() or the model(x) or the coordinate system?
Does anyone know what im doing wrong or knows how to do this?
I am working with the kinect and am placing the kinect point cloud in the 3d space using the original axis as my reference. now the kinect is placed at a angle somewhere in 3d space.
Currently i set up my scene and then pushmatrix() and translate/rotate to the kinect camera origin and then place the point cloud. then pop matrix and work with the data.. Now the kinect can calculate the centre of gravity for users but it is in the kinect 3d space cooridnates, and as the kinect uses the z axis as the "traditional" x axis everything is a bit strange. But my point cloud lines up and all is good.
The question: is there an easiy way to translate this point in 3d kinect space to the original reference space? im pretty sure i saw a native function somewhere that does this but i cant remember.so basically i need the centre of gravity in the original space
Im wondering if anyone can explain something to me or just let me know the "standard" way of doing this. Im new to 3D programming and i would like to get some advice on the whole navigating in 3d thing.. What i want to do is pretty simple, i want to place an object (just a line for now) on the origin of the 3d plane going up in the z direction. i then want to place the point cloud i have from a kinect in the 3d space. I know the offset from the origin in the real world as well as the rotations of the camera in the real world. no this is where im not to sure about how this is done normally.
As i understand so far. the camera is at the origin, and most programs i have seen translate the "subject they are wanting to see away from the origin. so the camera always stays stationary. i have seen the camera functions and havent had much success with it. what i thought of doing was to draw my "subject" at the origin, then pushmatrix() and translate to the Kinect origin and draw the point cloud then popmatrix() and set the camera to a position i want and point it at the origin. is this the right way? i have tried it and it doesnt seem to work.
Is it easier to leave the camera at (0,0,0,) and translate to my new origin and draw my "subject" then translate to the kinect origin and draw the point cloud?
I recently got a new PC and installed processing. My pc is 64 bit windows 7 and I seem to be having a problem with the java process not ending when I close my program after running it. if i click the X in the top corner my whole processing IDE freezes and unfreezes when i go in and end the java.exe * 32 process. Now if i click the stop button in the IDE it doesnt freeze but it doesnt end the java process either. So after a bit of testing i have tons of java.exe*32 processes running.
I have both java 32 bit and 64 bit installed. could this be the problem? any hel would be appreciated.
this is my 1st post. Im working on a project where i basically want to take an image that has three colours, separate the colours and then create an STL file of each colour regions.
Does anyone know of a simple way or to point me in the right direction. I can easily separate the colours, but am finding it hard to get started on the STL/mesh.
