SPEEDUP OF DETECTION AND ESTIMATION OF E. COLI CONTAMINATION USING AI-BASED OBJECT DETECTION

Problem: Detection of E.coli contamination and speedup the detection process, while providing quantifiable measure to estimate the growth level of contamination. Used Techniques: Convolutional Neural Network (CNN) supported by object detection modeling based on directions of the shape of the E.coli bacteria. Used Data: E.coli contamination images taken from Universe online repository and the milk is used as a food to be infected. Results: E. coli contamination poses a significant health risk, making its detection crucial. Leveraging advanced technologies like artificial intelligence (AI) and deep learning (DL), this study introduces an intelligent system designed to identify E. coli contamination in milk. The system employs object detection modeling to achieve higher accuracy in recognizing contaminants, while k-medoids clustering groups the detected objects. The cluster size serves as an indicator to estimate the contamination’s growth level. Compared to existing methods, the proposed system demonstrates superior performance, achieving 96% detection accuracy and a 7-second reduction in detection time. Keywords: E. coli, Object Detection, Clustering, Accuracy, Artificial Intelligence