Hough transform plays a very important role in computer vision as well as image processing. It is used for the detection of a known ob ject and the estimation of the object's orientation and location, when there are noises and distortions in the images. This paper first gives a brief introduction of Hough transform, and then proposes two statistical methods for solving the problem of estimating the shape parameters and also the problem of quantizing the parameter space. Finally a brief discussion of related statistical problems and further research directions.