Journal of Image Processing and Artificial Intelligence

Airline Twitter Sentiment Classification Using Machine Learning Techniques

Sujithra M, Kishore Kumar G — Fri, 12 Jan 2024 00:00:00 +0000

Today, there is more and more interest in social media. Numerous social media provide an outlet for the expression and spread of opinions about a broad range of topics, both publicly and privately. One of the social networks which have become popular is Twitter. Twitter provides organizations with a fast and effective way of analysing customers' views on the key to success in the marketplace. The sentiment analysis project is to assess the content of a given text's emotions, whether there are positive, negative or neutral feelings. Creating a vision test is a method used to evaluate people's opinions. This article shows how to create an opinion poll and how to remove many tweets. The results divide consumers' opinions of tweets into positive and negative. With the help of sentiment analysis, they will identify their weak points based on negative sentiments and start improving. Similarly, they will also identify strong points based on positive sentiments and they will try to maintain those with the same consistency or keep it up in future.

Assessing the Efficacy of Transfer Learning in Chest X-ray Image Classification for Respiratory Disease Diagnosis: Focus on COVID-19, Lung Opacity, and Viral Pneumonia

Chandrashekar Uppin, Gilbert George — Wed, 17 Jan 2024 00:00:00 +0000

Compared to alternatives like Polymerase Chain Reaction (PCR), the Chest X-Ray (CXR)-based method, which falls under Computer-Aided Diagnostic (CAD) approaches, provides a cost-effective solution for early-stage diagnosis of respiratory diseases, including Covid-19, Lung Opacity, and Viral Pneumonia. However, the utilization of CXR-based techniques for respiratory disease diagnosis has been relatively limited, with only a few studies exploring this approach. This research paper delves into the utilization of three distinct architectures—VGG16, VGG19, and MobileNet to classify three respiratory ailments mainly COVID-19, lung opacity, Viral Pneumonia and Normal healthy Lungs. The study incorporates a combination of transfer learning and a custom model, training models from the ground up. The methodology employed in this study centres around Computer-Aided Diagnosis (CAD) using Chest X-Rays (CXRs). Notably, this approach stands out as a cost-effective alternative in comparison to other diagnostic techniques such as Polymerase Chain Reaction (PCR), CT scans, and various medical procedures. Despite its cost-effectiveness, the adoption of CXR-based techniques for diagnosing respiratory diseases remains somewhat constrained. The research discussion section addresses this limitation by presenting and analysing the obtained results. We obtained an accuracy of 98% using the famous mobile Net architecture using Transfer learning, with VGG16 and VGG19 we obtained an accuracy of 97%.

Classification of Flowers Using Neural Networks Approach

Sandeep, Yogeesh. G. H, Shruthi, Divya. H. T, Leelavathy. T — Fri, 19 Jan 2024 00:00:00 +0000

This article proposed a technique for the classification of flowers based on texture and shape features using Feed-forward Neural Networks. This paper provides a unique solution for classifying flowers based on their texture, and geometrical value. The proposed method has three steps: (i) Segmentation; (ii) Masking; and (iii) classification. The segmentation was achieved by a familiar existing method called structural matrix decomposition (SMD) and considered two types of features called, GLCM (Gray-level co-occurrence matrix) and Shape features. In this proposed work, we adapted the Feed-forward NN algorithm for the classification of seven varieties of flower images. Experimentation was conducted using a dataset of 270 images of 7 classes to demonstrate the proposed model's performance. The experiment results demonstrate that a combination of GLCM, Hue-GLCM and Geometrical features gives a 96.90 % of accuracy rate. An experimental result shows the efficiency of the proposed approach.

The Review of Technologies for Unmanned Toll-Tax Collection By Means of Digital Image Processing

Pandit T. Yewale — Sat, 20 Jan 2024 00:00:00 +0000

Digital image processing has become an integral part of various industries as it helps in monitoring processes, detecting faults, ensuring quality control, and more. It is a specialized engineering science that deals with image enhancement, correction, feature extraction, surveillance, etc. Image processing is a complex and challenging field that requires a lot of experimentation and testing to find the right solution, while theoretical knowledge is essential, practical implementation is equally important. It involves a lot of algorithm revision, parameter estimation, and comparison of different solutions. Thus, choosing the right software development environment that is flexible, comprehensive, and well-documented is crucial. It can significantly affect the cost, development time, and portability of image processing solutions.

This project involves installing cameras and processing devices at every toll booth to capture the registration numbers of passing vehicles. The captured images will be processed using noise reduction techniques and contrast stretching to improve the clarity of the registration number. Using a text recognition algorithm, the registration number will be converted into a text file and sent to a remote server where it will be searched for in the database to debit the corresponding credit/debit card/bank account. The project uses the Optical Character Recognition (OCR) approach for text recognition and intends to use morphological operations to handle exceptional conditions.

A Brain Tumor Detection and Growth Rate Calculation from MRI Images Using Morphological and Watershed Operation

Tue, 06 Feb 2024 00:00:00 +0000

Advancements in imaging, particularly MRI technology, have ushered in a paradigm shift in brain tumor segmentation within medical science. Our cutting-edge system goes beyond conventional methods, ensuring meticulous examinations through detailed MRI imaging and the assignment of a unique patient ID for comprehensive record-keeping. This centralized identifier facilitates continuous tracking during subsequent evaluations, seamlessly integrating new MRI data into the patient's medical history. Employing state-of-the-art algorithms, the system conducts in-depth analyses, comparing current and historical tumor characteristics. This automated process aids in assessing improvements or worsening conditions, calculating tumor growth rates, and providing crucial insights for medical professionals. The continuous patient ID enables seamless comparisons, allowing for precise analysis and informed decision-making. This revolutionary approach significantly elevates the accuracy and efficiency of brain tumor diagnostics, heralding a new era in medical imaging. By empowering healthcare professionals with enhanced tools for comprehensive patient care, our system contributes to the ongoing evolution of precision medicine and advances the frontier of diagnostic capabilities in neurology.