Human species has been using their facial expressions and facial muscles in order to effectively communicate for millions of years now. Expression and emotions have been captured and stored into many forms since then like paintings, pictures, videos etc. Research is also being conducted in various forms in order to successfully identify the emotion of a person with help of technologies and methods such as knowledge based techniques, statistical methods and hybrid approaches. Recently artificial intelligence has proven its effectiveness in classifying or predicting the human emotion based on accumulated data in form of audio, video, text etc... This research proposes to design a system to classify the real-time emotions from numerous facial expressions and its features while using techniques such as Convolution Neural Network(CNN) algorithm. Along with CNN, this research also proposes to utilize various other tools and technologies such as OpenCV library, Tensorflow, Keras, Python, Visual Studio Code, etc.