Material flow characteristics for biodiesel

Abstract

The precise identification and characterization of small different oils are a necessary requirement for the completion of valuable screening. It is necessary to create an automated technique to identify the different types of oil. Many scholars have presented various methods in the past, but predicting with precision has always been a difficult issue. In this paper, an artificial neural network-based system for determining oil kinds is provided. Biodiesel is the one which can prepared naturally from either vegetable oils or animal fat. It is very difficult to identify the different oils, which are mixed together. In this paper, an automated testing technique is proposed based on the image processing techniques. Four different oil samples have been taken and they are classified using artificial neural network based algorithms. Results are promising to develop an automated tool. Wavelets are used in this work to decompose in to different frequency rang. Many types of wavelets are available for decomposing the images. Some types of wavelets suitable for medical images and some are for classification purpose. The wavelets are chosen based on the it’s characteristics like decomposition levels and vanishing moments. In this work wavelet transforms are used for decomposing purpose.The oils are classified based on the features using Artificial neural network.