This chapter introduces the Hough Transform, a fundamental technique for detecting geometric shapes by mapping image points into a parameter space and identifying parameter values that receive maximal voting. The traditional Hough Transform for line detection is first presented using the ρ–θ representation, which resolves the instability of vertical-line slopes in Cartesian form. Although robust against noise and discontinuities, the classical method suffers from high computational cost due to its exhaustive voting process. To address this limitation, an accelerated approach based on a known-point voting strategy is discussed, significantly reducing computation through coarse-to-fine slope estimation. Finally, the extension of Hough Transform to curve detection and the challenges associated with high-dimensional parameter spaces are examined.