Phenology is one effective way of tracking environmental changes through the study of plant’s periodic events and their relationship to climate. One promising research direction in this area relies on the use of vegetation images to track phenology changes over time. In this paper the authors propose the use of a new representation of time series to improve plants recognition rates in automatic image-based plant identification systems.
The technique, called recurrence plot, allows representing repeated events (the recurrence of states) on time series into two-dimensional representation. The features are extracted from this representation and then used as input to a learning method. The main contributions of the approach are: (i) an effectiveness analysis of the recurrence plots approach in different hours of day; (ii) a comparative study and correlation analysis between recurrence plots and visual rhythm approaches; and (iii) the adoption of a successful classifier fusion framework that combines the most suitable classifiers using both approaches.
Fabio A. Faria, Jurandy Almeida, Bruna Alberton, Leonor Patricia C. Morellato, Ricardo da S. Torres. Fusion of Time Series Representations for Plant Recognition in Phenology Studies. Pattern Recognition Letters. ISSN 0167-8655. doi:10.1016/j.patrec.2016.03.005