Automatic marking is mainly for digitizing surfaces. Photo4D can automatically mark a second image based on points marked in the first. This provides an effective way to mark and precisely correlate a large number of points in different images. First, users mark many points in one image, either manually or by using auto-detection. Then, the auto-marking feature can be invoked to find corresponding points in other images.
The Mark subpage in the Mark page of the Main Window is designed for this purpose.
Using auto-marking involves the following steps:
The new marks are all highlighted in green.
It is helpful to understand the basic idea of how Photo4D performs auto-marking. For each selected mark in the reference image Photo4D tries to find a corresponding mark in the working image in the following way. First, Photo4D estimates a mark's location in the working image. This estimation is based on the location of points that have been marked in both the reference image and working image. This is why users need to manually mark three or more points in the working image and merge them correctly to their corresponding points in the reference image.
Next, Photo4D uses a mask centered at the mark in the reference image to detect a similar feature point in the working image. The search range specifies the search area. It is a square centered at the estimated location with a boundary equal to plus and minus the search range value.
Finally, if the largest score of detected mark is lower then the score threshold, the auto-marking process stops. If two or more points overlap, all points except the first are deleted. This process is repeated until all the selected marks in the reference image are examined and transposed to the working image.
Again, auto-marking works well when the surface is smooth and feature points are well tagged. Otherwise, it may not work as desired. If it is totally inaccurate, user should use the traversal feature to manually mark the working image and then use the mark adjustment feature described in Section "Adjust Marks".
Auto-detection and Auto-marking are the major features which support surface digitizing. Auto-detection detects similar feature points in one image, while the auto-marking finds corresponding points in the other images. After marking the feature points, users may go through the steps to compute the (x,y,z) values for all the points.