Evaluating the Quality of Soft Computing in a Fog Computing Environment

The article deals with the problem of assessing the quality of soft computing on the example of solving the problem of processing video images. The factors influencing the accuracy of event detection based on video information analysis and the most popular algorithms used to determine them are indicated. The problem of comparing the results of experiments in quantitative terms (times, percentages) is indicated, which arises due to the heterogeneity of architectures, algorithms and means of implementing soft computing. A solution method is proposed. A system for evaluating the quality of soft computing has been developed and the effectiveness of its application has been proven on the example of processing video images from cameras installed in the subway. The quality of calculations was evaluated based on the accuracy of event detection and the speed of information processing. It was found that the data processing speed indicators for various soft computing methods differ by ±1.2 times, the accuracy values vary in the range of ±11.7%. The proposed method can be applied in other areas where it is required to evaluate the quality of soft computing, regardless of architectures, algorithms, and development tools.