摘要
足迹图像作为犯罪现场的重要痕迹物证之一,在破解串并案上有着不可忽视的作用.传统的足迹图像检索与匹配,需要耗费大量的时间与人力,极大地影响了破案进展.卷积神经网络(CNN)在图像识别与检索上表现出很好的效果.面向公安足迹图像比对实战需求,提出了一种基于卷积神经网络的足迹图像检索与匹配方法,对检索结果设置不同检索区,可以满足不同业务需求.初步实验表明该方法的有效性和实用性.
Footprint images, as one of the important evidences of crime scenes, can＇t be ignored in the cracking of serial cases. Traditional footprint comparison and retrieval require a lot of time and manpower,greatly affecting the progress of the case. Convolutional Neural Network（CNN） has shown good results in image recognition and retrieval. In order to meet the actual needs of public security footprint image retrieval, this paper proposes an approach to footprint image retrievaling and matching based on convolutional neural network,and sets different search areas for search results to meet different business requirements. Preliminary experiments show that the proposed approach is effective and practical.
作者
陈扬
曾诚
程成
邹恩岑
顾建伟
陆悠
奚雪峰
Chen Yang;Zeng Cheng;Cheng Cheng;Zou Encen;Gu Jianwei;Lu You;Xi Xuefeng(School of Electronic and Information Engineering,Suzhou University of Science and Technology,Suzhou 215009,China;Suzhou Key Laboratory of Virtual Reality and Intelligent Interaction,Suzhou University of Science and Technology,Suzhou 215009,China;Conunand Center of Kunshan Public Security Bureau,Suzhou 215300,China)
出处
《南京师范大学学报（工程技术版）》
CAS
2018年第3期39-45,共7页
Journal of Nanjing Normal University(Engineering and Technology Edition)
基金
国家自然科学基金(61750110534
61728205)
苏州市科技发展计划(重点实验室SZS201609/产业前瞻性项目SYG201707)
关键词
深度学习
卷积神经网络
足迹检索
图像处理
deep leaming
convolutional neural network
footprint searching
image processing
