Motion Analysis In Still Images

Date of Award

12-1-2023

Degree Name

M.C.S. (Master of Computer Science)

Department

Department of Computer Science

Advisor/Chair

Advisor: Tam Nguyen

Abstract

In the vast and intricate research area of computer vision, image classification has numerous real-world applications. Major study areas include the introduction of images in real-world applications like social entertainment, security, and healthcare. In this research, we provide a novel method for identifying optical illusions in images. Motion analysis is one of the means of describing various types of illusion images. For this proposed method we built a dataset and trained a neural network for classification. Using the dataset of 600 illusion images, the approach trains a complex deep neural network to investigate the effects of patterns, colors, and forms on visual perception. This network has a pre-trained model that is used to identify motion illusions in images. The images are mostly classified under motion illusion images and still non-illusion images. For the pre-trained model training, a comparative method was also utilized for state-of-the-art models. These models and mainly VGG16, ResNet-50, and a self-built Convolution Neural Network (CNN). We achieved good results on the above-mentioned training models. This research may eventually lead to the development of a new field of illusion detection study.

Keywords

MotionAnalysis, StillImages, ComputerVision, ImageClassification, OpticalIllusions, NeuralNetworks, VisualPerception, PatternRecognition, DeepLearning, VGG16 ResNet-50, ConvolutionalNeuralNetworks, IllusionDetection, ImageProcessing, PatternAnalysis, MachineLearning, ComparativeAnalysis, VisualIllusions, ImageRecognition, ComputationalModels, ArtificialIntelligence

Rights Statement

Copyright © 2023, author.

Link to Full Text

Share

COinS