Estimation of coniferous plants state using computer vision methods

Authors: Kataev M. Yu., Kislov A. V., Samohin E. A.

Annotation: Evaluating the ecology of an urban area is a complex process that is associated with numerous spatialtemporal measure-ments and the specifying of various parameters of soil and air. However, such measurements does not represent a pure monitoring, since the area of the average city is quite large and only a few points can be tracked on a regular basis. The article proposes an approach that is sensitive to changes that can occur in the of air and soil ecology and is associated with the study of the conifers’ state. Conifers are green all year round, planted in many parts of the city and their RGB images are available on a regular basis. Each image can be segmented using Greenness indexes and then the state of the plants estimated from the analysis. The condition of the coniferous plant is associated with a color change (from green to yellow). A software system using the proposed methods and algorithms is described. The processing results of real measurement data showing the operability of the proposed approach is presented.

Keywords: coniferous plant, software, clustering, vegetation index, ecology