TY - JOUR
T1 - Understanding the Hough transform: Hough cell support and its utilization
AU - Hansen, Klaus
AU - Andersen, Jens Damgaard
PY - 1997
Y1 - 1997
N2 - The Standard Hough Transform (SHT) is used to find lines, circles and other image features in edge maps containing edge points. Edge points which potentially can ‘vote' for a given line (called its ‘line support') are highly dependent of line position and orientation due to the boundedness of images. If accumulator array peaks are used to find lines then peripherally located lines are missed. Uniform feature detection probability in all parts of the image requires normalisation of the accumulator array by the support. The paper describes some further applications of the concept of support when finding lines (e.g. determining dominant directions), and gives some guidelines on the choice of the coordinate system origin.
AB - The Standard Hough Transform (SHT) is used to find lines, circles and other image features in edge maps containing edge points. Edge points which potentially can ‘vote' for a given line (called its ‘line support') are highly dependent of line position and orientation due to the boundedness of images. If accumulator array peaks are used to find lines then peripherally located lines are missed. Uniform feature detection probability in all parts of the image requires normalisation of the accumulator array by the support. The paper describes some further applications of the concept of support when finding lines (e.g. determining dominant directions), and gives some guidelines on the choice of the coordinate system origin.
KW - Faculty of Science
KW - Hough transform; Curve detection; Hough space cell support; Bias; Parameter quantisation
U2 - 10.1016/S0262-8856(96)01128-6
DO - 10.1016/S0262-8856(96)01128-6
M3 - Journal article
SN - 0262-8856
VL - 15
SP - 205
EP - 218
JO - Image and Vision Computing
JF - Image and Vision Computing
IS - 3
ER -