ABSTRACT
This paper a new proposes Algorithm to overcome
the drawbacks of the generalized Hough transform,
namely its computational complexity and storage
requirement. For decreasing the Computation time,
the algorithm performs the Hough transform by (1)
decomposing an image into small blocks, (2)
estimating line parameters through least-squares line
fitting for each block, and (3) removing detected
lines while performing the Hough transform for the
remaining blocks. For reducing the memory storage,
the proposed algorithm utilizes a data structure, list
to represent accumulators. The method significantly
reduces the computational complexity and storage
required to evaluate all possible parameters without
using a accumulator array. Experimental results are
given to show the correctness and effectiveness of the
proposed method.