Das et al. [1] proposed an indexing method to index flower patent images using domain knowledge. The flower was
segmented using an iterative segmentation algorithm with domain knowledge driven feedback. In their work, the image
color is mapped to names using the ISCC-NBS color system and X Window system. Each flower image is discretized in HSV
color space, and each point on the discretized HSV space is mapped to a color name in the ISCC-NBS and X Window systems in
order to index the flowers. Yoshioka et al. [6] performed a quantitative evaluation of petal colors using principal component
analysis. They considered the first five principal components (PCs) of a maximum square on the petals. The quantitative
evaluation indicates that the different PCs correspond to different color features of petals such as color depth, difference in
color depth in upper and lower parts of the image, etc.