Self-Configuring CVS to Discriminate Rocket Leaves According to Cultivation Practices and to Correctly Attribute Visual Quality Level

Computer Vision Systems (CVS) represent a contactless and non-destructive tool to evaluate monitor the quality of fruits vegetables. This research paper proposes an innovative CVS, using Random Forest model automatically select relevant features for classification, thereby avoiding their choice through cumbersome error-prone work human designers. Moreover, three color correction techniques were evaluated compared, in terms classification performance identify best solution provide consistent measurements. The proposed CVS was applied fresh-cut rocket, produced under greenhouse soilless cultivation conditions differing irrigation management strategy fertilization level. first aim this study objectively estimate levels (QL) occurring during storage. second non-destructively, manner, approach digital images obtained product. achieved accuracy about 95% QL assessment 65–70% discrimination approach.