• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Microprocessors and microsystems >Simulation of sports action picture recognition based on FPGA and convolutional neural network
【24h】

Simulation of sports action picture recognition based on FPGA and convolutional neural network

机译：基于FPGA和卷积神经网络的体育动作图像识别模拟

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The actions of the opposition are basically communicated. In addition, it searches for relevant information about competitor photo detection and sports activity photos that can be implemented in external photos. Judging from the situation in the military, the situation is going well. Therefore, the evaluation development diagram is to consider the purpose of image confirmation development. Video detection involves the entire system. The structural configuration depends on the hardware, including the Field Programmable Gate Arrays (FPGAs). In this proposed combines several image processing methods to produce a clear difference image, while at the same time being more robust to changes in light intensity and background scenes. Edge-unquestionable check appraisal, grayscale orchestrating, target acquirement, target attestation, picture district headway. For a long time, it has set up the structure for displays to test the relevance of investigation frameworks. This licenses you to see competitors' affirmations, game statements, and sports clues, to say the least. The video detection results from the relevant surveys indicate that a solution exists. Use Convolutional Neural Network (CNN) instead of encoding immutable features to classify images into human and non-human categories and actions.
机译：反对派的行为基本沟通。此外，它还搜索有关竞争对手照片检测和体育活动照片的相关信息，可以在外部照片中实现。从军队的情况判断，情况进展顺利。因此，评估开发图是考虑图像确认发展的目的。视频检测涉及整个系统。结构配置取决于硬件，包括现场可编程门阵列（FPGA）。在此提出中，组合了多种图像处理方法来产生清晰的差异图像，而同时对光强度和背景场景的变化更加稳健。边缘毫无疑问的检查评估，灰度协调，目标获取，目标证明，图片区前往。很长一段时间，它已经设置了显示器的结构，以测试调查框架的相关性。这是您最少地说的竞争对手的肯定，游戏声明和体育线索。来自相关调查的视频检测结果表明存在解决方案。使用卷积神经网络（CNN）而不是编码不可变特征，将图像分类为人类和非人类类别和行动。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号