Difference between point and dense clouds

[agluck]

As a newbie I don't understand why there are a different number of points in the point cloud and dense cloud. What does each represent and what occurs when the dense cloud is created from the point cloud? Thanks.

[bgroff]

Someone will correct me if I'm wrong, but the 'point cloud' is just the image tie points. I think Pix4D may call this the Ray Cloud. The 'dense cloud' is truly the point cloud (e.g. las file).

[James]

yes the sparse point cloud points are tie points. each point corresponds to a feature that was identified in more than one photo, and deemed to be a valid match.

dense cloud points are subsequently calculated by rectifying image pairs such that epipolar lines become parallel (the transformation is dependent on camera orientation parameters that were derived from the tie points) and then a different algorithm can be used to match pixels to pixels in these image pairs.

i think...

[agluck]

I greatly appreciate the replies, particularly the technical explanation of dense cloud points by James....but I don't know enough about computer terms to understand.  What does is mean "by rectifying image pairs such that epipolar lines become parallel"? How does this create the "Points" in the dense cloud, when the terminology seems to indicate that what is created are "lines", not points?

[Kiesel]

@agluck,

why there are a different number of points in the point cloud and dense cloud?

Because they are results of different steps in workflow/ processes. The sparse point cloud is a result of the align step, where tie points are used and at first distortion parameters and camera positions are unknown. With the help of the Tie points both is then calculated. With known distortion parameters and camera position it is possible to generate stereo pairs and calculate the dense point cloud from them in a second wokflow step. This is was is described by James: 

by rectifying image pairs such that epipolar lines become parallel (the transformation is dependent on camera orientation parameters that were derived from the tie points) and then a different algorithm can be used to match pixels to pixels in these image pairs.

.