首页> 外国专利> COLLISION INJURY PREDICTION MODEL CREATION METHOD, COLLISION INJURY PREDICTION METHOD, COLLISION INJURY PREDICTION SYSTEM, AND ADVANCED AUTOMATIC ACCIDENT NOTIFICATION SYSTEM

COLLISION INJURY PREDICTION MODEL CREATION METHOD, COLLISION INJURY PREDICTION METHOD, COLLISION INJURY PREDICTION SYSTEM, AND ADVANCED AUTOMATIC ACCIDENT NOTIFICATION SYSTEM

机译:碰撞伤害预测模型创建方法,碰撞伤害预测方法,碰撞伤害预测系统和高级自动事故通知系统

摘要

To improve the prediction accuracy of a collision injury level of an injured object person at the time of an automobile collision accident.SOLUTION: A collision injury prediction model creation method for creating a collision injury prediction model used to predict a collision injury level of an injured object person at the time of automobile accident occurrence includes the following steps of: creating a database composed of collision accident images of injured object people observed at the times of automobile accidents occurred in the past and injury information; creating a large number of learning datasets in which collision accident image data is associated with injury information data by converting the database into a format suitable for implementation of machine learning; performing deep learning for the learning datasets by selecting an artificial intelligence model using a multilayered structure algorithm modeling a human brain neural circuit; and creating a learned artificial intelligence model having undergone deep learning as the collision injury prediction model.SELECTED DRAWING: Figure 4
机译:解决方案:一种碰撞伤害预测模型创建方法,用于创建用于预测受伤者的碰撞伤害水平的碰撞伤害预测模型,以提高汽车碰撞事故时受伤对象人的碰撞伤害水平的预测准确性。发生汽车事故时的对象人包括以下步骤:创建一个数据库,该数据库由过去发生汽车事故时观察到的受伤对象人的碰撞事故图像和伤害信息组成;通过将数据库转换为适合实施机器学习的格式,创建大量学习数据集,其中碰撞事故图像数据与伤害信息数据相关联;通过使用模拟人脑神经回路的多层结构算法选择人工智能模型,对学习数据集进行深度学习;并创建经过深度学习的学习型人工智能模型作为碰撞伤害预测模型。选定的图纸:图4

著录项

  • 公开/公告号JP2020061088A

    专利类型

  • 公开/公告日2020-04-16

    原文格式PDF

  • 申请/专利权人 JAPAN AUTOMOBILE RESEARCH INSTITUTE;

    申请/专利号JP20180193564

  • 发明设计人 KUNITOMI SHOHEI;KONOSU ATSUHIRO;

    申请日2018-10-12

  • 分类号G16H50/70;G08G1;G08B25/04;G08B25/08;G08B25;G06Q10/04;G06N20;G06N3/08;B60R21;

  • 国家 JP

  • 入库时间 2022-08-21 11:36:38

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

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