Scholarlink

When using remote sensing pictures for marine security, ship detection is essential. The deep learning method for identifying ships from satellite photos is covered in this research. In order to achieve integrity hashing is included. This model makes use of a supervised method for classifying images, and then use YOLOv3 for object recognition, feature extraction from Deep CNN. Using class labels, semantic segmentation and picture segmentation are used to determine the object category of each pixel next with the satellite image’s bounding box is defined and helps us to identify the position of ship and ship count. We have implemented hashing with the help of SHA-512. It is used to encode the ship count and locations. A dataset of around 2,30,000 photos from Kaggle ship detection is used for the proposed model. Thirty percent of the data is used for testing, while the remaining seventy percent is used for training. The bounding box location and the ship count are the input data used by the hash algorithm. In order to achieve security, we use SHA-512 which maintains security for the transmission of data.

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 Ship Detection Accuracy in Satellite Images with YOLOv3 and Convolutional Feature Extraction

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Karunakumar Valicharla,
  • Rama Krishna Eluri,
  • Moturi Sireesha,
  • S. N. Tirumala Rao

摘要

When using remote sensing pictures for marine security, ship detection is essential. The deep learning method for identifying ships from satellite photos is covered in this research. In order to achieve integrity hashing is included. This model makes use of a supervised method for classifying images, and then use YOLOv3 for object recognition, feature extraction from Deep CNN. Using class labels, semantic segmentation and picture segmentation are used to determine the object category of each pixel next with the satellite image’s bounding box is defined and helps us to identify the position of ship and ship count. We have implemented hashing with the help of SHA-512. It is used to encode the ship count and locations. A dataset of around 2,30,000 photos from Kaggle ship detection is used for the proposed model. Thirty percent of the data is used for testing, while the remaining seventy percent is used for training. The bounding box location and the ship count are the input data used by the hash algorithm. In order to achieve security, we use SHA-512 which maintains security for the transmission of data.