SBIR Phase I: Flexible and High Performance Biometric Tools for Small-to-Medium Scale Identification Applications

SBIR 第一阶段：适用于中小型识别应用的灵活高性能生物识别工具

• 批准号: 0636227
• 负责人: Chiman Kwan
• 金额: $ 10万
• 依托单位: Signal Processing, Inc.
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-01-01 至 2007-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0636227&HistoricalAwards=false
• 关键词: SBIR; Phase; Flexible; Performance; Biometric

This Small Business Innovation Research (SBIR) Phase I research project proposes a library of low-cost, flexible, and high performance biometric tools that target the small- and medium-sized business market. These tools, comprising both hardware -- low-cost fingerprint device and microphone, and associated software, can work individually or in a combined manner via multi-modal fusion. For low cost fingerprint devices, these novel image processing tools can robustly and reliably capture fingerprints for recognition. In particular, for the low cost sweep sensor, algorithms are proposed to splice fingerprints obtained from multiple sweeps into a high quality fingerprint. Additionally, the project will develop low-cost touch sensor, with novel algorithms to mosaic multiple fingerprints from a single user to form a single complete and high quality fingerprint. For voice print recognition, the research will enhance the recognition performance by combining an ordinary microphone with a glottal motion sensor that can help the speaker verification process in two ways. The combined market for fingerprint and voice will be 2.7 billion dollars by 2008. There are many potential users of these flexible and low-cost biometric tools. For applications involving computer network security, the use of fingerprint and voice prints, these biometric tools will enhance existing password techniques. Other potential applications include community and building entrance checkpoints.

该小型企业创新研究 (SBIR) 第一阶段研究项目提出了一个针对中小型企业市场的低成本、灵活且高性能的生物识别工具库。这些工具由硬件（低成本指纹设备和麦克风）以及相关软件组成，可以单独工作，也可以通过多模式融合以组合方式工作。对于低成本指纹设备，这些新颖的图像处理工具可以稳健可靠地捕获指纹进行识别。特别是，对于低成本扫描传感器，提出了将多次扫描获得的指纹拼接成高质量指纹的算法。此外，该项目将开发低成本触摸传感器，采用新颖的算法来拼接单个用户的多个指纹，形成单个完整且高质量的指纹。对于声纹识别，该研究将通过将普通麦克风与声门运动传感器相结合来增强识别性能，声门运动传感器可以通过两种方式帮助说话人验证过程。到 2008 年，指纹和语音的综合市场将达到 27 亿美元。这些灵活且低成本的生物识别工具有许多潜在用户。对于涉及计算机网络安全的应用，指纹和声纹的使用，这些生物识别工具将增强现有的密码技术。 其他潜在的应用包括社区和建筑物入口检查站。