Image processing based edibility analysis of spinach leaves using machine learning approach

Abstract

Analyzing the edibility of food consumed by the human body is very crucial to identify the nutritional values absorbed. Lack of the right amount of nutrients can lead to various health issues like food poisoning, low immunity and nutritional diseases. Thus, identifying such problems at the stage of consumption can help in preventing several foodborne diseases and improve health. But this aspect is given little importance in our country, due to the heavy expenses involved and the infeasibility of large scale deployment of existing methods, which are mainly chemical experiments. Thus, the main goal of this work is to provide a simpler, cost-effective solution to address the given issue. Green leafy vegetables, specifically spinach plants are considered for this research as they are highly nutritious with very low longevity. Given the normal storage conditions, the shelf life of spinach leaves can be extended to a maximum of 5-7 days[1]. During the course of this research, we analyze the edibility of spinach leaves using Image Processing techniques and Machine Learning in order to provide simpler solutions that can replace the existing methods. A data-set was created to capture the deteriorating stages of the spinach leaves at regular intervals of time for ten days. Image Processing techniques were used to extract the chlorophyll and nitrogen content of the leaves. By using Machine Learning, these values were correlated with the age of the leaf. After the training process, testing was performed to identify the performance of the proposed system.