Application of Haar Cascade Classifier for Kitchen Safety Monitoring

The purpose of this study aims to provide kitchen safety using Haar Cascade Classifier and alert systems. This kitchen safety monitoring study uses the Haar cascade for detecting humans inside the kitchen. The study uses the DHT22 temperature sensor to monitor the temperature of the kitchen, IR Flame sensors to monitor the stove activity, and ESP 32 cam to provide live surveillance. The blynk application displays the parameters for user monitoring. The alert system consists of two alerts; The initial alert activates when no human is present in the kitchen, yet the stove is active for 40 seconds. The initial alert consists of short pulses from the buzzer. The Final alert activates when no human is present, yet the stove is active for 70 seconds. The Final alert consists of a continuous buzz and an SMS message alerting the user. The alert system also alerts the user when the temperature exceeds the threshold. The alert system achieves a 10/10 reliability during the test if the stove is left unattended and a 93.25% accuracy for the classifier.