SBIR Phase I: Intelligent Cameras for Crime Fighting

SBIR 第一阶段：用于打击犯罪的智能摄像机

• 批准号: 9560676
• 负责人: ChiMan Kwan
• 金额: $ 7.47万
• 依托单位: Intelligent Automation, Inc
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1996
• 资助国家: 美国
• 起止时间: 1996-04-01 至 1996-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9560676&HistoricalAwards=false
• 关键词: SBIR; Phase; Intelligent; Cameras; Crime

This Small Business Innovation Research Phase I project proposal describes the next generation of security cameras. These cameras will provide two improvements over current technology. First, the `Smart Camera` will provide much higher resolution images of person's in its field of view than is possible with today's security cameras. With today's security cameras, a suspect's image captured by a security camera could easily have only a small number of pixels actually focused on the person. This may not be enough for identification or for irrefutable evidence in court. The `Smart Camera` will capture a much higher resolution image. Based on initial computations, for typical situations, thousands of fixed cameras with telephoto lenses would be required to reliably capture the same resolution image of a suspect as would be captured by a single `Smart Camera.` The second improvement over current security cameras is that this type of camera captures images with high precision 3 dimensional information. If the camera were mounted two meters from a person as the person entered a bank, Government building, etc., the face data, including depth information across every feature of the face, would be captured with an accuracy of approximately .4 mm. Accurately measured three dimensional images of a person's face may prove to be nearly as unique as a fingerprint. `Faceprints` could be captured visibly as a subject walks into a building or even walks along a street. Since 3 dimensional images contain literally an entirely new dimension of data, images should be able to be compared more quickly and reliably. Images of people entering major facilities such as train or bus stations could be automatically matched in real-time against most wanted lists, or suspects in recent crimes. The bombing of the Federal Building in Oklahoma City has focused attention on the need for increased security. Potential criminals should be significantly deterred by better security cameras, and by automatic means to identify and apprehend criminals. For those that are not deterred, the system will provide a next generation of tools for identification and apprehension.

这份小型企业创新研究第一阶段项目提案描述了下一代安全摄像头。这些相机将提供两个改进，比目前的技术。首先，“智能摄像机”将提供比今天的安全摄像机更高分辨率的人的图像。使用今天的安全摄像头，安全摄像头捕获的嫌疑人图像很容易只有少量像素实际聚焦在该人身上。这可能不足以在法庭上识别或作为不可辩驳的证据。“智能相机”将捕捉更高分辨率的图像。根据初步计算，对于典型情况，需要数千个带有长焦镜头的固定摄像头才能可靠地捕获与单个“智能摄像头”捕获的分辨率相同的嫌疑人图像。 与当前安全摄像机相比的第二个改进是，这种类型的摄像机捕获具有高精度三维信息的图像。如果当人进入银行、政府大楼等时摄像机安装在离人两米处，将以大约0.4mm的精度捕获包括面部的每个特征的深度信息的面部数据。精确测量的人脸的三维图像可以证明几乎与指纹一样独特。当一个人走进一座建筑物，甚至沿着一条街道走的时候，“脸印”就可以被明显地捕捉到。由于三维图像包含了一个全新的数据维度，因此图像应该能够更快速、更可靠地进行比较。人们进入火车站或汽车站等主要设施的图像可以自动实时与最受欢迎的通缉名单或最近犯罪的嫌疑人进行匹配。 俄克拉荷马州市联邦大楼爆炸事件使人们集中注意加强安全的必要性。应该通过更好的安全摄像头以及自动识别和逮捕罪犯的手段来大大阻止潜在的罪犯。对于那些没有被吓倒的人，该系统将提供下一代识别和逮捕工具。