Structure-from-motion-python icon indicating copy to clipboard operation
Structure-from-motion-python copied to clipboard

Implementation based on SFMedu Princeton COS429: Computer Vision http://vision.princeton.edu/courses/SFMedu/ but on python + numpy

Results 7 Structure-from-motion-python issues
Sort by recently updated
recently updated
newest added

I wonder if I have multiple fixed-angle videos, how can I use them to implement SFM algorithm?

What's the meaning of variable name`str`? Why choose a name could be easily confused with string as convention. https://github.com/aferral/Structure-from-motion-python/blob/374e2d13e0372ece081cfcd2e96a47664bfb3962/utils/graph.py#L60

Traceback (most recent call last): File "main.py", line 110, in graphList[i] = triangulateGraph(graphList[i],imageSize) File "/content/drive/SFM/utils/graph.py", line 84, in triangulateGraph x[cnt,:] = newGraph.obsVal[newGraph.ObsIdx[i][ind],:] IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis...

line 79, in mergeG merged.obsVal = np.vstack([merged.obsVal, gB.obsVal[bObbsIdx, :]]) IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices help me

File "main.py", line 62, in imageSize = list(map(int,listaImages[0].shape)) TypeError: 'map' object is not subscriptable

It seems that the point clouds from different graphs have not been transformed and integrated properly.

stuck after feature point detection ? What should i do now ?