In the 21st century, stress has emerged as a major epidemic, impacting various sectors. The current methods to assess stress and related mental health issues are still mostly based on self-reporting questionnaires, which are time-consuming, prone to bias, do not allow for continuous monitoring, and are not scalable. This results in mental health issues being diagnosed only after the symptoms are severe enough to be noticed by others. Poor mental health is linked to a prolonged state of negative emotions. The significance of these conditions has led researchers to explore the use of emotion recognition systems to assess mental health issues. This chapter offers a look at the state of the art surrounding emotion, encompassing its definition, fundamental theories, and various measurement methodologies. Our review aims to elucidate how emotion recognition technologies can be achieved, towards enhancing mental health interventions. Poor mental health not only diminishes the well-being of individuals but also elevates the risk of errors and accidents, thereby compromising operational effectiveness.