Rendering images and voxelizing the images

1

I am using the shapenet dataset. From this dataset, I have 3d models in .obj format. I rendered the images of these 3d models using pyrender library which gives me an image like this : enter image description here

Now I am using raycasting to voxelize this image. The voxel model I get is something like below :

enter image description here

I am not able to understand why I am getting the white or light brown colored artifacts in the boundary of the object.

The reason I could come up with was maybe the pixels at the boundary of the object contain two colors, so when I traverse the image as numpy array, I get an average of these two colors which gives me these artifacts. But I am not sure if this is the correct reason.

If anyone has any idea about what could be the reason, please let me know

mascot

Posted 2020-01-06T16:15:53.370

Reputation: 19

2Is that related to AI? Please explain – None – 2020-01-06T18:04:54.657

1This is off topic. @nbro – Clement Hui – 2020-01-07T00:40:55.197

I checked on google for the forums where I can ask questions about computer vision. This forum came up citing the reason that since "computer-vision" is available as one of the tags, questions related to this can be posted here – mascot – 2020-01-08T11:16:19.700

Please refer to the help center for questions that is on topic. https://ai.stackexchange.com/help/on-topic

– Clement Hui – 2020-01-08T15:04:39.137

The computer vision tag in this case means this: For questions related to computer vision, which is an interdisciplinary scientific field (which can e.g. use image processing techniques) that deals with how computers can be made to gain high-level understanding from digital images or videos. For example, image recognition (that is, the identification of the type of objects in an image) is a computer vision problem This is defined in the tag wiki of the computer-vision tag. Your question clearly does not fit the description of the tag. – Clement Hui – 2020-01-08T15:05:59.427

@mascot If you can put this problem in the context of computer vision, then this question may be on-topic here. Maybe try reformulating your question. – nbro – 2020-01-08T15:15:26.617

No answers