Detection of Deer in Images by Computer Vision Methods

IF 0.5 Q4 PHYSICS, MULTIDISCIPLINARY

Optoelectronics Instrumentation and Data Processing Pub Date : 2024-06-20 DOI:10.3103/s8756699024700109

S. N. Tereshchenko, A. L. Osipov

引用次数: 0

Abstract

An approach to applying machine learning methods for automatic detection of deer individuals in images has been studied. Neural network technology has been used to accurately count the number of deer from photographs. Deep learning methods for convolutional neural networks (ResNet 50, DenseNet, CenterNet, Inception V3, and Xception) were used in conjunction with the transfer learning technique. Based on the faster R-CNN Resnet50 network, a neural network was trained to identify deer individuals from graphic images with an accuracy of 0.91 on a sample using the F1-score metric with a threshold value of 0.6.

Abstract Image

查看原文本刊更多论文

用计算机视觉方法检测图像中的鹿

摘要研究了一种应用机器学习方法自动检测图像中鹿个体的方法。神经网络技术被用于从照片中精确计算鹿的数量。卷积神经网络（ResNet 50、DenseNet、CenterNet、Inception V3 和 Xception）的深度学习方法与迁移学习技术结合使用。以速度更快的 R-CNN Resnet50 网络为基础，对神经网络进行了训练，使用阈值为 0.6 的 F1 分数指标，从图形图像中识别鹿个体的准确率达到 0.91。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Optoelectronics Instrumentation and Data Processing PHYSICS, MULTIDISCIPLINARY-

CiteScore

1.00

自引率

50.00%

发文量

期刊介绍： The scope of Optoelectronics, Instrumentation and Data Processing encompasses, but is not restricted to, the following areas: analysis and synthesis of signals and images; artificial intelligence methods; automated measurement systems; physicotechnical foundations of micro- and optoelectronics; optical information technologies; systems and components; modelling in physicotechnical research; laser physics applications; computer networks and data transmission systems. The journal publishes original papers, reviews, and short communications in order to provide the widest possible coverage of latest research and development in its chosen field.