The paper presents five algorithms of XLog language of HTML and XML documents processing. They are the following ones: the Lucas–Kanade algorithm, processing patterns, mean shift, Viola–Jones object detection, and continuously adaptable shift. The comparative analysis of these algorithms is conducted; their advantages and disadvantages are revealed. Their efficiency is evaluated. The results of each algorithm performance are presented with regard to the problem of XLog language of HTML and XML documents processing. The authors check the algorithms for probable errors of the 1st and 2nd types, which can occur when using the algorithms. They also analyze the algorithm performance peculiarities in video stream processing. Two most appropriate algorithms for XLog language of HTML and XML documents processing are chosen. For the analyzed problem to be solved, the authors introduce an integral algorithm, using capabilities of both the Lucas–Kanade algorithm and Viola–Jones object detection.