In behavioural biometrics, one common method of authentication is offline signature recognition with forgery detection. An innovative offline signature forgery detection method based on Hough transform characteristics is presented in this research. One method for extracting features is the hough transform, which may identify regular forms like lines, curves, ellipses, and so on. Following the preprocessing procedures for the signature samples, we use the Hough transform to extract the shape characteristics. We do this by extracting attributes from ten real signatures and five fake signatures belonging to twenty different individuals. After that, the correlation coefficient and the thresholding approach are used to identify forgeries. A 92% accuracy rate was achieved using the suggested approach.