Neuro Quantology | September 2022 | Volume 20 | Issue 9 | Page 4638-4650 | doi: 10.14704/nq.2022.20.9.NQ44537 Puneeth Kumar B S, Mr S.V.Ramanan, Galiveeti Poornima, Mrs. Jayashree M Kudari, Dr. C M Velu, Hari Krishna silamanthula/ Image Analysis using Machine Learning and Traditional Image Processing Approaches 4638 Image Analysis using Machine Learning and Traditional Image Processing Approaches Puneeth Kumar B S *1 , Mr S.V.Ramanan 2 , Galiveeti Poornima 3 , Mrs. Jayashree M Kudari 4 , Dr. C M Velu 5 , Hari Krishna silamanthula 6 *1 Assistant Professor, Department of Computer Science, St. Joseph's University, Langford road, Bengaluru, Karnataka, India. 2 Assistant Professor, Department of Electronics and Communication Engineering, PPG Institute of Technology, Coimbatore, Tamil Nadu, India. 3 Assistant Professor, Department of Computer Science and Engineering, Presidency University, Bangalore, Karnataka, India. 4 Associate Professor, School of CS and IT, Jain Deemed to be university, Bangalore, Karnataka, India. 5 Professor, Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha University, SIMATS, Chennai, Tamilnadu , India. 6 Assistant Professor, Department of Fine Arts, K L E F Deemed to be University, Guntur, Andhra Pradesh, India. Abstract The demand for speedier algorithms that can extract information from photos is becoming more crucial due to the continuously growing amount of image data in practically all disciplines. The practise of utilising digital image processing methods to extract significant information from a 2-dimensional picture is known as image analysis. The term "digital image processing" refers to a group of methods used in pre- processing (such as noise removal and image enhancement), image compression, feature extraction (such as edges and contours), and various key point detection methods (such as corners and joints, areas with particular colours or textures). This study focuses on the recognition and representation of texture, a crucial aspect of picture analysis. Utilizing a well-known texture identification technique built on filter banks, we conduct tests on a variety of artificial and real-world photos. We pinpoint certain particular situations in which the algorithm fails and suggest a change to the original approach that produces better segmentations. Keywords: Machine Learning, Digital Image, Filters, Feature extraction and Noise . DOI Number: 10.14704/nq.2022.20.9.NQ44537 Neuro Quantology 2022; 20(9):4638-4650 1. Introduction The practise of utilising digital image processing methods to extract significant information from a 2-dimensional picture is known as image analysis. The term "digital image processing" refers to a group of methods used for preprocessing (such as noise removal and image enhancement), image compression, feature extraction (such as edges and contours), and various key point detection methods (such as corners and joints, areas with particular colours or textures) [1]. The processes involved in image analysis now include the following thanks to the development of machine learning (ML) and deep learning (DL) methods, which are the major topics of this paper: Choose a machine learning (ML) algorithm that is appropriate for the task at hand, such as Support Vector Machine (SVM), Naive Bayes (NB), Expectation Maximization (EM), or k-means, or a suitable deep learning architecture, such as Convolutional Neural Network (CNN), Long Short Term Memory (LSTM), or Autoencoder. The task at hand, the training data that are available, and the computational resources that are available all play a role in selecting the DL architecture or ML method [2]. Preprocessing and feature