Punchcard: VideoTrace

VideoTrace creates low polygon count models.  The number of polygons in a model relates directly to the number of polygons you draw.  That means that it’s most efficient to model objects with large planar surfaces.  Curved objects can easily be modelled, however, by breaking them up into segments.  That’s what is going to happen in […]

One of the advantages of having the user control the modelling process in VideoTrace is that it is possible to model parts of an object which don’t actually appear in the input image set.  See the following example of modelling the Sydney Opera House; the video shows only one side of the object, but both […]

This example shows the process of modelling a house using VideoTrace

and the corresponding 3D model

This clip shows more of the modelling process.  The result is a texture mapped model of a workstation.

This example shows the process of generating a model of an archway.  The process interesting because the extrusion tool is used to model parts of the object not visible in the image sequence to be modelled.  The following shows the final model

and this the result of inserting the model into Google Earth

This video shows the […]

The Sydney Opera House presents a number of challenges to any image-based modelling package, but the most significant is probably the fact that the ’sails’ are complex curves with practically no texture.  This makes it very difficult for automated methods to identify the shape accurately.  The user interaction in VideoTrace overcomes this problem, however, as […]