Fruit Freshness Detecting System Using Deep Learning and Raspberry PI

Food Quality Checking is an essential thing to get rid of food poisoning and diseases, Monitoring the decayable products, and deterioration of fruit at their early stage will decrease the loss of fruit wastage and ensure freshness. Recent technological advancements like Artificial Intelligence and Machine Learning (ML) has important methodologies which may improve the fruit quality cautionary process’s value and time potency. Machine vision techniques are currently wide to discover the quality of fruits. Image process is typically the primary step in detecting the standard of fruits. the method starts by capturing the image of the fruits using camera available. Then, the image is transmitted to the process stage wherever it will extract the fruits options like form, size and color. These processes are done by the image process. It helps to spot and compare the fruit form, size and color with the trained datasets which can be done throughout the prepare and testing stage. This paper aims mainly for blind people where it is difficult for them to check whether the fruit is consumable or not . In this project, a recommendation system is build using various Deep Learning algorithms like Visual Geometry Group (VGG)16 and Convolutional Neural Network(CNN) that is employed for feature Extraction and machine learning algorithms like Logistic Regression, Light Gradient Boosting(LGB) and Random Forest which are used for prediction and a comparative study is made, the performance of each algorithm is calculated which is then integrated with raspberry pi, so that when a customer places the fruit inside a box available, the camera which is present will capture the fruit from all side and will detect whether it is a fresh or rotten fruit and the output is told to the user via voice command/Speaker.