Approaches to assess the development and implementation of computer vision technologies in the agro-food complex based on the patent landscape and agent modeling

Abstract

The paper presents approaches to the analysis and evaluation of digital technologies with agent-based modeling, scientific and patent landscaping. We tested these approaches on the example of computer vision systems in the agro-food sector. The result of the tests is a collection of full-text documents (more than 1.1K patents and 3.5K scientific papers) that makes it possible to assess the level of publication and patent activity of international research on computer vision for the period 2010–2021. We also proposed an approach to build an agent-based model that evaluates the effects of the use of computer vision. The results can be applied to create state programs for digitalization, considering global trends to ensure national food security.