Scholarlink

<p>Image classification is a crucial task in computer vision that involves identifying and labeling objects within images. This study proposes a hybrid feature extraction method for image classification, combining local features (SIFT, Haralick descriptors) with deep features (VGG19) to capture a wide range of image characteristics. We use k-means clustering and locality-preserving projection for feature selection and dimensionality reduction. The method was tested on the Caltech-101 dataset and achieved high accuracies of 96.18%, 96.76%, 97.39%, and 97.78% using k-NN, MLP, Random Forest, and Adaptive Boosting classifiers, respectively. Cross-validation results confirm the model’s robustness, making it suitable for real-world applications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Handcrafted Feature and Deep Features Based Image Classification Using Machine Learning Models

  • 期刊论文

  • 发表日期：

    2025年2月5日

查看全文
  • Anupam Yadav,
  • Ali Khatibi,
  • H. S. Shreenidhi,
  • Saroj Kumar Gupta,
  • Abhilasha Jadhav,
  • Mandeep Kaur Chohan,
  • G. Sanyasi Raju,
  • Ahmed Alkhayyat

摘要

Image classification is a crucial task in computer vision that involves identifying and labeling objects within images. This study proposes a hybrid feature extraction method for image classification, combining local features (SIFT, Haralick descriptors) with deep features (VGG19) to capture a wide range of image characteristics. We use k-means clustering and locality-preserving projection for feature selection and dimensionality reduction. The method was tested on the Caltech-101 dataset and achieved high accuracies of 96.18%, 96.76%, 97.39%, and 97.78% using k-NN, MLP, Random Forest, and Adaptive Boosting classifiers, respectively. Cross-validation results confirm the model’s robustness, making it suitable for real-world applications.