Abstract:Image classification is the process of finding common features in images from various classes and applying them to categorize and label them. The main problem of the image classification process is the abundance of images, the high complexity of the data, and the shortage of labeled data, presenting the key obstacles in image classification. The cornerstone of image classification is evaluating the convolutional features retrieved from deep learning models and training them with machine learning classifiers. This study proposes a new approach of “hybrid learning” by combining deep learning with machine learning for image classification based on convolutional feature extraction using the VGG-16 deep learning model and seven classifiers. A hybrid supervised learning system that takes advantage of rich intermediate features extracted from deep learning compared to traditional feature extraction to boost classification accuracy and parameters is suggested. They provide the same set of characteristics to discover and verify which classifier yields the best classification with our new proposed approach of “hybrid learning.” To achieve this, the performance of classifiers was assessed depending on a genuine dataset that was taken by our camera system. The simulation results show that the support vector machine (SVM) has a mean square error of 0.011, a total accuracy ratio of 98.80%, and an F1 score of 0.99. Moreover, the results show that the LR classifier has a mean square error of 0.035 and a total ratio of 96.42%, and an F1 score of 0.96 comes in the second place. The ANN classifier has a mean square error of 0.047 and a total ratio of 95.23%, and an F1 score of 0.94 comes in the third place. Furthermore, RF, WKNN, DT, and NB with a mean square error and an F1 score advance to the next stage with accuracy ratios of 91.66%, 90.47%, 79.76%, and 75%, respectively. As a result, the main contribution is the enhancement of the classification performance parameters with images of varying brightness and clarity using the proposed hybrid learning approach.

Various Image Classification Using Certain Exchangeable Image File Format (EXIF) Metadata of Images

Detecting Image Tampering Using Feature Fusion

Authentication of Digital Image using Exif Metadata and Decoding Properties

Modeling The Exif-Image Correlation For Image Manipulation Detection

Boosting Image Classification with LDA-based Feature Combination for Digital Photograph Management

EXIF as Language: Learning Cross-Modal Associations Between Images and Camera Metadata

ON IMAGE CLASSIFICATION: CITY IMAGES VS. LANDSCAPES

A New Proposed Hybrid Learning Approach with Features for Extraction of Image Classification

Content-Based Image Retrieval and Feature Extraction: A Comprehensive Review

Analysis of errors in exif metadata on mobile devices

Fast Feature Extraction Based on Multi-feature Classification for Color Image.

Exif2Vec: A Framework to Ascertain Untrustworthy Crowdsourced Images Using Metadata

Estimating EXIF Parameters Based on Noise Features for Image Manipulation Detection

Recovery of descriptive information in images from digital libraries by means of EXIF metadata

Detecting Interesting Regions in Photographs --- How Metadata Can Help

Digital Image Tamper Detection Techniques - A Comprehensive Study

@ A novel framework for image categorization and automatic annotation (English)

Classification of digital photos taken by photographers or home users

A Knowledge Enforcement Network-Based Approach for Classifying a Photographer's Images

Simulation of Computer Image Recognition Technology Based on Image Feature Extraction

Intelligent photography software information processing system design based on data analysis