Computer Vision―The Frontier of Modern Environmental Diagnostics: A Review

Abstract

Computer vision (CV), in combination with various sensors and image analysis algorithms, is a frontier direction in diagnosing the state of the environment and its biogenic and abiogenic objects. The work generalizes scientific achievements and identifies scientific and technical problems in this area of research based on the conceptual system of analysis on the time axis: from implemented achievements as part of the past and present to original new solutions—the future. Our work gives an idea of three areas of application of CV in diagnosing the state of the environment: phenotype recognition in digital images, monitoring of living and abiogenic objects, and development of new methods for identifying pollution and its consequences. The advantages of CV, which can be attributed to scientific achievements in this field of research, are shown: an increase in the volume of analyzed samples, simultaneous analysis of several parameters of the object of observation, and leveling of subjective evaluation factors. The main CV problems currently solved are the accuracy of diagnostics and changing quality of the survey, identification of the object of analysis with minimal operator participation, simultaneous monitoring of objects of different quality, and development of software and hardware systems with CV. A promising direction for the future is to combine the capabilities of CV and artificial intelligence. Thus, the review can be useful for specialists in environmental sciences and scientists working in interdisciplinary fields.