基于多模态人机交互的汽车天窗控制系统设计-学报杂志社

期刊信息

合肥工业大学（自然科学版），2025年11月，第48卷第11期：1448-1451，1458

DOI: 10.3969/j.issn.1003-5060.2025.11.002

作者信息

方博一 $ ^{1} $，汪春华 $ ^{2} $，耿展 $ ^{2} $，姚卫军 $ ^{2} $，江昊 $ ^{3} $，陈奇 $ ^{1} $

(1. 合肥工业大学机械工程学院, 安徽合肥 230009; 2. 黄山奥特斯电气股份有限公司, 安徽黄山 245400; 3. 合肥工业大学汽车与交通工程学院, 安徽合肥 230009)

摘要和关键词

摘要: 随着汽车科技的发展,智能座舱利用人机交互技术能够为驾乘人员提供更舒适的体验,然而目前的人机交互主要依赖基于摄像头的手势识别技术,存在成本高、算力需求大的问题。为此,文章设计了基于红外手势识别和语音识别的人机交互方案。红外手势识别方案以STM32F103为控制核心,设计2种实现方式,即采用APDS-9960-3.3 RGB传感器的集中式和使用通用红外传感器的分布式;语音识别方案采用US516P6纯离线语音识别芯片作为核心;通过搭建控制模块并与汽车天窗连接,对2种方案的性能进行测试。试验结果表明,所设计的方案在有效降低成本、减少算力消耗的同时,仍能保持较高的识别准确率。该研究为汽车智能座舱的发展提供了一种新的技术路径。

关键词: 手势识别；红外传感器；语音识别；人机交互；智能座舱

Authors

FANG Boyi $ ^{1} $, WANG Chunhua $ ^{2} $, GENG Zhan $ ^{2} $, YAO Weijun $ ^{2} $, JIANG Hao $ ^{3} $, CHEN Qi $ ^{1} $

(1. School of Mechanical Engineering, Hefei University of Technology, Hefei 230009, China; 2. Huangshan Automaster Electric Co., Ltd., Huangshan 245400, China; 3. School of Automobile and Traffic Engineering, Hefei University of Technology, Hefei 230009, China)

Abstract and Keywords

Abstract: With the development of automotive technology, intelligent cockpits using human-machine interaction techniques can provide a more comfortable experience for drivers and passengers. However, current mainstream human-machine interaction techniques rely on camera-based gesture recognition technology, which involves high costs and significant computational power requirements. Therefore, this paper proposes a human-machine interaction scheme based on infrared gesture recognition and voice recognition. The infrared gesture recognition scheme uses the STM32F103 as the control core, and two infrared gesture recognition methods, centralized and distributed, are designed using the APDS-9960-3.3 RGB sensor and general infrared sensors, respectively. The voice recognition scheme adopts the US516P6 pure offline voice recognition chip as the core. A control module is built and connected to the vehicle sunroof to test the performance of each scheme. It is verified that the designed system effectively reduces costs and computational power consumption while maintaining the accuracy of human-machine interaction recognition. This provides a new technical path for the development of intelligent cockpits in vehicles.

Keywords: gesture recognition; infrared sensor; voice recognition; human-machine interaction; intelligent cockpit

基金信息

安徽省重点研究与开发计划资助项目（2022a05020024）；合肥工业大学校地合作产业创新引导资金资助项目（JZ2023YDZJO011）

合肥工业大学学报(自然科学版)

导航菜单

基于多模态人机交互的汽车天窗控制系统设计