Real-Time Quality Assurance of Fruits and Vegetables using Hybrid Distance based MKELM Approach

Sustainable development relies on a number of pillars, one of which being agriculture. Sustainable agriculture, in light of expected population expansion, must ensure food security while remaining economically and socially viable and having a minimal impact on biodiversity and natural ecosystems. Deep learning has shown to be an advanced method for analyzing large amounts of data, having applications in fields as diverse as image processing and object recognition. Recently, it’s being used in the fields of food engineering and science. Food recognition, quality detection of produce, meat, and seafood, the food supply chain, and contamination were only some of the problems these systems set out to solve. Artificial intelligence (AI) is a common tool in the field of precision agriculture for making predictions about the quality of harvested crops. This is especially true when assessing crops at various post-harvest stages. Certain postharvest diseases or damages, like rot, can completely wipe out crops and even produce toxins that are hazardous to humans, making disease and damage identification a top priority. Preprocessing with a gabor filter, enhancement with HE, segmentation with a K-means algorithm, and feature extraction with LBP and BIC make up the suggested method. Lastly, DB-KELM is used to train the model. As compared to ELM and KELM, the proposed method performs better.