Flower Classification Supported by Digital Imaging Techniques

Abstract

The application of imaging techniques in agricultural process is becoming a very promising tool in supporting pattern development as well as shape and size survey. The analysis of individual elements in the classification process or at the initial specie recognition meets timing and precision requirements. Brazil exhibits an outstanding potential to develop flower and ornamental plants production for the domestic as well as for the international markets. Nowadays flower production provides an income close to US$ 1.3 billion/year, gathering more than 500 producers concentrated in an area of 8500 hectares, generating an income of U$ 35 million in the year of 2007. The Brazilian flowers and ornamental plants exportation strategy has the objective of orienting and directing the specific potentiality of flower production to take the commercial opportunities offered by the international market, emphasizing post-harvest technology. Based on what it has been explained above, the objectives of this research work can be summarized as to test the application of digital imaging on gerbera (Gerbera jasmesonii) stem. A total of 120 stems digital images were processed through the ‘ImageJ’ free software. From generated stem images, computer routines were written able to recognize and to classify gerbera according to flower opening, stem straightness as well as peduncle inclination.