Abstract: At least two cameras (10,12) view two points (A,B) displaced by a known distance on an article (S). Using a known relationship between two points (A,B) and the known position of cameras (10,12) in space the location of the viewed points (A,B) in space are determined by calculation. Similarly, the coordinates of other points on to article (S) are determined and hence dimensions such as width are determined.

Abstract: A method and apparatus for providing information as to the cross-sectional shape or profile of a body is disclosed in which a known datum in the form of a bar 50 having a cross-shaped cross-section is detected by a camera 14. A body is then detected by the camera 14 to provide preliminary information as to the shape or profile of the body and the preliminary information is transformed in accordance with a variation of the detected datum from the known datum to provide a better indication of the true cross-sectional shape or profile of the body. The known datum 50 enables at least three non-colinear reference locations to be defined and calibration parameters are determined from the reference locations so that exact reference locations can be determined.