This paper presents a model of evaluation for computing performance and usability benchmarks. The target evaluation is context analysis of video streams. The paper describes the required system components, system architecture and evaluation model, as well as the test cases that can be performed using the designed environment. During the modelling, design, implementation and preliminary tests of the environment, some issues, problems and good-to-know key points were discovered which are explained in this paper. Reference guidance for future test cases is also given. The modelling and experimental results presented in this paper are based on the test environment, and the observations of the cooperation project between Tampere University of Technology Pori Unit and Keio University Shonan Fujisawa Campus.