This research article highlights the innovations in the area of non-invasive methods for the quality prediction of Edible oils. The concept of quality is dependent on its related characteristics which are used for evaluation of the edible oils. Currently existing non-invasive methods include spectroscopy, Nuclear Magnetic Resonance (NMR), and hyper-spectral imaging methods. Prediction of the oil quality depends on the colour of the oil. A quality detection scheme with colour extraction technique from the images of various cooking oils is developed through image analysis that uses the pixel intensities in Red (R), Green (G) and Blue (B) plane. A brief outline about the commercialization for quality prediction, control and monitoring of edible or cooking oils is discussed here. The various types of oils used for quality prediction include highly virgin olive oil (HVOL), Light Olive Oil (LOL), Sunflower Oil (SO), Canola oil (CAO), Grape Seed Oil (GSO), Corn Oil (COO), Avocado Oil (AO), Peanut Oil (PO), Palm Oil (PAO) and Sesame Oil (SO). The images of these oils are subjected to preprocessing, attribute extraction and finally the prediction of the quality using Convolutional Neural Network (CNN). The conventional imaging methods require upgradation in terms of data gathering, processing time, maintenance cost reduction and enhancement in market value. On the whole, this research article enlightens that the non-invasive techniques have the scope for horticultural application and can be used to determine its quality