Data driven modelling for banana ripeness assessment

Banana is a widely grown and extremely perishable fruit. Thus post-harvest quality changes quickly. The main reason for this study is to determine the quality of banana and segregate them using low cost method. Conventional methods include sorting and grading based on human evaluation, firmness check by penetrometer. These are non-effective to give uniform and quick results. In this study the bananas are classified based on dielectric spectroscopy and image processing techniques. The dielectric spectroscopy applied is low cost, non-destructive, quick and easy to operate. The ethylene sensor is a low cost capacitive based sensor. The result is achieved with fuzzy logic controller. These method used gives uniform and quick results.