• Content count

  • Joined

  • Last visited

Community Reputation

120 Neutral

About shkumar07

  • Rank
  1. I have to render an .obj 3d model on top of a video captured from a webcam or a depth cam. Whatwould be the best way to do that . Right now the way I am trying to do this is rendering the 3d model on an offscreen buffer using FBO , read pixels from it into an RGB array and sending it to a OpenCV window which is responsible for rendering the webcam video. The opencv window blends this pixel buffer and the video based on the deph of each pixel. Is there another better way ? Can opengl window receive the webcam feed and and render it as a texture with some part of the texture occluded by the 3d model. Another related question is , if i have to render a 3d model using opengl window, i can use glut or win32 window and render dirctly in to the window using opengl's draw APIs OR i can render in offscreen buffer, read the pixel data and render it as an image, say using opencv. Both do the job but what is the difference.
  2. Per Vertex color to Texture

    I understand that a mesh can be rendered just based on vertex color, however we were having problems loading a .obj file with per vertex color ( without texture map) in Unity game engine. There were a couple of other reasons that we need to bake vertex color in to texture programmatically.  
  3. Per Vertex color to Texture

    Thanks Hodgeman, Jason . However the problem is that mesh doesn't have texture coordinates .
  4. I have a .obj mesh which has per vertex color information stored in it. There are no texture associated with this mesh. Does someone know a method/source code/algorithm which can help me generate a texture image from per vertex color ? 
  5. So does incompatible pixel descriptor cause depth test issue. As the attacked screenshots show - the first image shows how the rendering should look, but the second image is how it actually gets drawn on remote desktop. My application has three different Opengl windows with different context. The first window renders correctly during remote desktop connection , but the other two render as shown in image 2 of the attached image. Why doesn't incompatible pixel format affect the first window.
  6. I am using OpenGL 1.1 for rendering. On local machine it works fine but when I remote desktop I see a couple of issue - 1. initially the drawing window is black and blank , i have resize it a little bit to render something on it. 2. sometimes surfaces show through each other, or surface further from the camera drawl shows up but not the surface which is close to camera at the same X,Y value. It seems depth test is failing. I get the same effect when I turn off hardware acceleration of the monitor. I understand that windows turns off hardware acceleration on remote desktops. Is there a work around for this ? Why does turning off hardware acceleration messes up the depth test and repainting of window
  7. Hi I am searching for a good third party library for 3D Data Compression and Decimation for mesh objects. If anyone in this group has a first hand experience about using external libraries for mesh compression and would wan to recommend it then please let me know. I need to use the library to save the compressed data on disk and then uncompress them when loading them in to memory. Thanks Shailesh