Computer vision methods and algorithms for embedded vision applications

用于嵌入式视觉应用的计算机视觉方法和算法

• 批准号: 484341-2015
• 负责人: Laganière, Robert
• 金额: $ 3.74万
• 依托单位: University of Ottawa
• 依托单位国家: 加拿大
• 项目类别: Collaborative Research and Development Grants
• 财政年份: 2017
• 资助国家: 加拿大
• 起止时间: 2017-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=618779
• 关键词: Computer; vision; methods; algorithms; embedded

The objective of this research is to develop practical computer vision algorithms to be used in embedded vision applications. This research will be done in partnership with Cognivue, an Ottawa-based company, specializing in designing programmable Image Cognition Processors (ICPs) that enable real-time embedded image and video analytics in smart device applications. The research will focus on identifying, developing, implementing and testing the most appropriate algorithms for specific computer vision applications.The identified field of applications are Stereovision, Augmented reality and object recognition, Face analysis and recognition. Once a particular algorithm is developed, the researchers will work in close collaboration with the company's engineers to deploy it on the Cognivue ICP board. This project represents an opportunity for Cognivue to test, validate and improve its APEX CV Application Interface (API) and architecture in the context of embedded vision applications. For the researchers of our lab, it will be an opportunity to deploy vision algorithms developed by the team of researchers in real-world applications.

本研究的目的是开发可用于嵌入式视觉应用的实用计算机视觉算法。这项研究将与Cognivue合作完成，Cognivue是一家总部位于渥太华的公司，专门设计可编程图像认知处理器（icp），使智能设备应用中的实时嵌入式图像和视频分析成为可能。该研究将集中于识别、开发、实施和测试最适合特定计算机视觉应用的算法。确定的应用领域是立体视觉，增强现实和物体识别，人脸分析和识别。一旦开发出特定的算法，研究人员将与公司的工程师密切合作，将其部署在Cognivue ICP板上。该项目为Cognivue提供了在嵌入式视觉应用环境中测试、验证和改进其APEX CV应用程序接口（API）和架构的机会。对于我们实验室的研究人员来说，这将是一个将研究团队开发的视觉算法部署到实际应用中的机会。