Detecting Stress in Software Professionals: A Machine Learning and Image Processing Approach

Stress can be defined as a state of worry or mental tension caused by a difficult situation. Stress is high in software profession because of their nature of work, target, achievements, night shift, over work load. This may lead to disease, chronic backache, headache, high BP, insomnia etc. No person can continue under stress for too long. By timely detection, it is possible to detect stress and take necessary action to overcome it. This paper suggests a method to detect stress using machine learning and image processing techniques. There are various stress detection systems available today. All stress detection systems have 3 basic steps: image preprocessing, Feature selection and classification. In this paper some of the methods are analysed. In the proposed methodology the image is preprocessed using gain and bias parameters. PCA is used as feature selection. The features are then used in the automatic classifier such as SVM for the automatic classification