Abstract:

Objective In order to improve the detection, recording and recognition ability of the optical imaging equipment in the public security audio-visual technology, and to solve the problem that the traditional two-dimensional camera causes the lack of information of the target object in the image and video, this paper introduces the multi-spectral light field camera into the field of public security audio-visual technology, which can simultaneously obtain the space, direction and spectral information of the object. The object characteristic information can be recorded more comprehensively. Methods In this paper, the functions of multi-spectral light field camera, such as full focus, super depth of field, three-dimensional imaging, viewpoint free transformation and multi-physical information snapshot imaging, are used in various fields of public security audio-visual technology, such as scene photography, material evidence inspection photography, criminal portrait, video investigation technology and police intelligent identification. Results The light field camera can effectively solve the photography problem of difficult material evidence such as potential material evidence out of focus; The preservation dimension of the image is upgraded to four dimensions, that is, 3D spatial information and spectral information, which greatly improves the feature points of material evidence and face recognition, and effectively improves the accuracy of material evidence comparison and the accuracy of portrait identification; In the application of police intelligent recognition technology, it can improve the accuracy of human behavior characteristics similarity quantification, as well as increase the success rate of iris acquisition and improving the accuracy of iris recognition. Conclusion Multi-spectral light field imaging technology can comprehensively improve the technical indicators in various fields of public security audio-visual technology, greatly improve the efficiency of discovery and identification of physical evidence on the scene, improve the integrity of evidence information preservation, improve the accuracy of portrait comparison in video investigation technology, and improve the success rate and accuracy of identification in police intelligent identification.

Key words:

public security audio-visual technology, multi-spectral, optical field imaging

摘要：

目的 为提升公安视听技术中光学成像设备对目标客体的发现、记录与识别能力，解决使用传统二维相机造成图像和视频中目标客体信息缺失的缺点，将多光谱光场相机引入公安视听技术领域，该相机可以同时获取被摄物的空间、方向信息及光谱信息，可以更加全面地记录客体特征信息。方法 将多光谱光场相机的全焦点、超景深、三维成像、视点自由变换、多物理信息快照式成像等功能，分别应用于公安视听技术的各个领域：现场摄影、物证检验摄影、刑事相貌、视频侦查技术、警务智能识别。结果 光场相机可以有效解决潜在物证失焦等疑难物证的拍摄问题；将影像的保存维度提升至四维，即三维空间与光谱维度，大幅提高物证与人脸识别时的特征点，有效提高物证比对精度与人像同一认定精度；在警务智能识别技术的应用中，可以提高人体行为特征相似度量化的精确性，以及调高虹膜采集成功率、提升虹膜识别精度。结论 多光谱光场成像技术可全面提升公安视听技术中各领域的技术指标，极大提高现场物证的发现与识别效率、提升物证信息保存的完整度，并且可提高视频侦查技术中人像比对精度，以及在警务智能识别应用中的识别成功率和准确性。

关键词:

 , 公安视听技术, 多光谱, 光场成像