pork adulteration in e nose dataset(7 levels of adulteration)
Description
Abstract Electronic noses (E-noses) are increasingly utilized for food authentication and adulteration detection. In the context of halal compliance, pork adulteration in beef remains a significant concern. This study presents an E-nose system developed using eight MQ series gas sensors (MQ2, MQ3, MQ4, MQ5, MQ135, MQ136, MQ137, and MQ138) alongside a DHT22 sensor for temperature and humidity monitoring. These sensors were integrated with an ESP32-WROOM microcontroller, with each sensor connected to an analog input pin. The system transmits raw analog-to-digital converter (ADC) values wirelessly to a web server hosted at iotwebserver.com. The web server is configured to hold a maximum of 50 data entries, after which older entries are overwritten. To preserve the full dataset, raw readings are periodically downloaded and merged after every 50 updates. Instructions: Fresh ground beef and pork were procured from the same source on the same date. A total of seven sample compositions (each 250 g) were prepared: 250 g beef (100% beef)-class 1 225 g beef + 25 g pork (90% beef, 10% pork)-class 2 175 g beef + 75 g pork (70% beef, 30% pork)-class 3 125 g beef + 125 g pork (50% beef, 50% pork)-class 4 75 g beef + 175 g pork (30% beef, 70% pork)-class 5 25 g beef + 225 g pork (10% beef, 90% pork)-class 6 250 g pork (100% pork)-class 7 The procedure for each combination was as follows: Sensors were preheated for 20 minutes before data acquisition. The prepared sample was placed in the E-nose sampling chamber. Sensor outputs were recorded every 1 minute for a duration of 124 minutes per sample. After each trial, the chamber was ventilated using a fan for 4 minutes to prevent residual gas interference. The dataset comprises exclusively raw ADC values. Each sample's data was saved in a .csv format, labeled according to its composition. Each record contains readings from the eight MQ sensors, followed by temperature and humidity values.