Surveillance is the monitoring of behavior. Furthermore, system surveillance is the process of monitoring the behavior of people, objects, or processes within systems for compliance with expected or desired norms in reliable systems for safety control (Cohen and Medioni, 1999). Video surveillance systems existed 25 years ago, when it started with a 100% analog system and gradually became a digital system. Closed circuit television (CCTV) camera is the most popular video surveillance due to its reliability and low price. The camera does not transmit images but records them, so that the user can always check what happened while he was away. It is widely used in public spaces and residences for security purposes. In many countries the use of surveillance videos has grown intensively. As a result, a large number of crimes have been successfully recorded by surveillance videos. Therefore, the question of identifying the person on recording media or digital images is more frequently brought to the attention of forensic experts. However, this can be a daunting task especially when the recordings are of poor quality and the enhanced footage failed to reveal the face (identity) of the suspect. Most of this problem is caused by system resolution, uncalibrated equipment, and the lack of an efficient and practical method to identify the criminal suspect through surveillance images. There are many private companies or public spaces that still use CCTV (analogue) in low resolution formats for surveillance purposes. Analog CCTV has a major disadvantage in terms of resolution compared to megapixel or high resolution CCTV (Cavoukian, 2001). According to Ratcliffe (2006) analog resolution is sufficient for generalized and generalized line and dot detection in images and for the use of intelligent markers in reference video recordings. On the other hand, to address the second (ii) problem, i.e. the uncertainties about the suspect's pose, shoes, hair and caps, several practical solutions are chosen, such as exclusively using images in which the perpetrator is feet, and report body length estimates that include shoes, hair, and caps. Some researchers use computer models of humans superimposed on images. The pose of these models can be changed to make them look like the perpetrator in the images. One of these approaches was published by (Lynnerup and Vedel, 2005). The most important conclusion is that for each case, reference video recordings of people with known lengths are crucial for estimating the systematic and random error.