VARIABILITY AND INTERDEPENDENCIES BETWEEN FRACTAL GEOMETRY PARAMETERS IN GRAPEVINE LEAVES

The grapevine leaves was analyzed based on fractal geometry, and the study evaluated interdependencies between parameters specific to fractal geometry. The grape varieties considered in the study were represented by Chasselas (Ch), Cabernet Sauvignon (CS), Muscat Hamburg (MH), Muscat Iantarnîi (MI), Muscat Ottonel (MO), Perlă de Csaba (PC), and Victoria (Vi). Fresh leaves were taken from each variety. The leaves were herborized to preserve the integrity of the leaves, and were scanned. Imaging analysis and fractal analysis were used to generate fractal geometry parameters: foreground pixels (FP), fractal dimensions D1 (MEE = 0.5), D2 (MEE = 0.2), BW parameter (black and white pixels), B+BW parameter (black + black and white pixels), W+BW parameter (white + black and white pixels). The descriptive statistical analysis generated the analysis report based on a number of 15 samples for each grapevine variety. Correlation analysis was used to evaluate the interdependence between fractal geometry parameters, for each grape variety. Correlations with different levels of intensity were recorded between the parameters of the fractal geometry, in relation to the vine variety. The values ​​of the correlation coefficient showed some matches and differences of the parameters of the fractal geometry in the seven grape varieties studied. In all cases, a very strong positive correlation was recorded between D1 and D2. Also, a negative correlation was recorded between W+BW and B+BW parameters, but of variable intensity.