On Choosing the Precedent Size in the Problem of Objects Detecting in Digital Images

Abstract

In the paper we consider the problem arising in the video surveillance systems development of choosing the precedent size when detecting objects in images, taking into account the optoelectronic system characteristics. The example shows that in order to solve the objects detection problem at different distances from the observer on digital images, based on comparison with precedents, it is essential to choose the appropriate precedent size, which depends on the digital video camera characteristics. Examples of applications in surveillance tasks of video cameras with different equivalent focal lengths are considered. The relations for determining the equivalent focal length of the lens (equivalent to 35 mm) are given based on the value of the actual focal length of the camera lens determined by its design features. Relations are given for calculating the precedent size corresponding to an observed object of a given size and located at a known distance from the observer, taking into account the characteristics of the optoelectronic system. Examples of object size values (pixels) in the image are given depending on the distance to the object of specified dimensions (m) on the observed scene at different values of the digital video camera characteristics. An algorithm has been developed for solving the problem of detecting objects in an image based on the analysis of precedents, which dimensions depend on the video camera characteristics, as well as on the desired object size and the distance from it to the observer. Examples of object detection in images are given, taking into account the digital video camera characteristics.