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

Enhancing Fruit Recognition with YOLO v7: A Comparative Analysis Against YOLO v4

  • Anand Pandey,
  • Shivansh Kumar,
  • Aaryansh Verma,
  • Shaurya Baijal,
  • Chiranjit Dutta,
  • Tanupriya Choudhury,
  • Jagdish Chandra Patni

摘要

This study uses YOLO v7 deep learning to recognize and localize fruits in complicated pictures. The detection effectiveness of YOLO v7 is compared to YOLO v4, a dominating fruit identification model. The YOLO v7 algorithm enhances fruit detection accuracy, as shown by increased average precision and other evaluation parameters. This study found that the algorithm might help manage planting activities and identify difficult situations. Fruit recognition and categorization in agriculture are stressed, as are deep learning models. The correct classification of fruits in complex environments remains difficult. The suggested solution uses YOLO v7’s unique features to identify and locate fruits in complex surroundings. A comparative analysis showed that YOLO v7 outperformed v4 in fruit recognition and classification. The study classifies fruits into 16 distinct categories and examines the possible uses of the approach for different kinds of agricultural products. Integrating supplementary datasets can enhance the deep learning algorithm’s ability to differentiate and classify diverse agricultural commodities, thereby improving crop management strategies and streamlining processes such as automated sorting, quality assurance, and inventory supervision. The findings of this research facilitate the creation of fruit categorization technologies that operate in real-time across diverse settings, thereby conferring advantages to enterprises in the fruit industry. The aforementioned technologies can increase efficiency, improve productivity, optimize resource allocation, and reduce losses resulting from spoilage and infections. Implementing these technologies mitigates potential harm to essential goods, thereby safeguarding the quality of produce and the well-being of consumers. The results of this study make a valuable contribution to the progress of real-time fruit classification technologies. This has practical implications for enterprises in the fruit industry, as it can enhance efficiency, generate financial gains, and safeguard commodities. The ramifications of these results hold great importance for the triumph and endurance of the industry.