语音合成研究及在智能监控中的应用开发毕业论文
2021-11-11 20:43:19
论文总字数:26127字
摘 要
智能监控系统是未来智能生活的重要部分,将语音合成与传统监控通过计算机技术相结合,是预防危害发生以及及时通知报警的重要手段。
在传统的监控系统中,往往需要大量的人工对视频进行多次的存储回放操作,这对于服务器的性能提出了非常高的要求,不利于监控工作的进行,并且卒然更换设备也会造成不小的浪费。因此在不对旧设备进行大规模改造的基础上,对监控系统加以改进,使之实现对监控画面的智能处理与自主分析,减轻人工压力,并且通过语音合成技术实现智能的语音报警,弥补人工监督下可能产生的疏忽,将监控的结果实时反馈给用户,从而实现自动化、智能化的监控系统。
本文主要介绍了智能监控系统的设计方案。智能监控系统主要分为七个模块:画面采集、图像识别、图像匹配、语音合成、信息上报、信息存储、网络监控,主要完成工作有:
(1)画面采集模块。实现了摄像头数据采集,将摄像头采集到的画面显示在监控框中,并同时进行录像与截图操作,其中截图供后续模块操作。
(2)图像识别模块。实现了对人脸的识别与框定,对采集到的监控画面的截图进行人脸识别,判断画面中是否有人存在,并产生报警文字信息。
(3)图像匹配模块。实现了两张图片的匹配,对给出的图像进行特征点提取,然后对两张图的特征点进行比较,如果相似度高则视为一致。
(4)语音合成模块。实现了根据文字对语音的合成,根据图像识别后产生的文字报警信息,智能的合成语音报警音频文件,再通过播放音频文件进行语音报警。
(5)信息上报模块。实现了报警信息由电脑程序向网页以及安卓客户端的推送,确保用户打开网页或客户端能实时收到报警信息。
(6)信息存储模块。实现了对监控采集信息以及图像处理结果的存储,通过数据库间接存储实时监控截图以及识别出的人脸信息。
(7)网络监控模块。实现了通过打开网页或安卓客户端,远程进行实时监控。
关键词:智能监控;语音合成;人脸识别;图像匹配;网络监控
Abstract
Intelligent monitoring system is an important part of intelligent life in the future. The combination of voice synthesis and traditional monitoring through computer technology is an important means to prevent the occurrence of hazards and timely notify the police.
In the traditional monitoring system, a large number of manual video storage and playback operations are often needed, which puts forward very high requirements for the performance of the server, which is not conducive to the monitoring work, and the sudden replacement of equipment will also cause a lot of waste. Therefore, based on the large-scale transformation of the old equipment, the monitoring system is improved to realize the intelligent processing and independent analysis of the monitoring picture, reduce the artificial pressure, and realize the intelligent voice alarm through the voice synthesis technology, make up for the negligence that can be generated under the artificial supervision, and feed back the monitoring results to the user in real time, so as to realize automation and intelligence The monitoring system.
This thesis mainly introduces the design of intelligent monitoring system. The intelligent monitoring system is mainly divided into seven modules: image acquisition, image recognition, image matching, speech synthesis, information reporting, information storage, network monitoring. The main tasks are as follows:
(1) Picture acquisition module. The camera data acquisition is realized, the image collected by the camera is displayed in the monitoring frame, and the video recording and screenshot operation are carried out at the same time, and the screenshot is for the subsequent module operation.
(2) Image recognition module. The system realizes face recognition and framing, and carries out face recognition for the screenshot of the collected monitoring screen, judges whether there is anyone in the screen, and generates alarm text information.
(3) Image matching module. The matching of the two images is realized, the feature points of the given image are extracted, and then the feature points of the two images are compared. If the similarity is high, it is regarded as consistent.
(4) Speech synthesis module. According to the text to speech synthesis, according to the text alarm information generated after image recognition, the intelligent synthetic voice alarm audio file is realized, and then the voice alarm is carried out by playing the audio file.
(5) Information reporting module. It realizes the push of alarm information from computer program to web page and Android client to ensure that users can open web page or client can receive alarm information in real time.
(6) Information storage module. It realizes the storage of monitoring acquisition information and image processing results, and indirectly stores the real-time monitoring screenshots and the recognized face information through the database.
(7) Network monitoring module. It realizes remote real-time monitoring by opening web page or Android client.
Key Words: intelligent monitoring; speech synthesis; face recognition; image matching; network monitoring
目 录
第1章 绪论 1
1.1 研究背景与意义 1
1.2 国内外研究现状 1
1.2.1 国内研究现状 1
1.2.2 国外研究现状 2
1.3 设计内容 3
1.4 论文章节安排 4
第2章 关键技术与系统实现 5
2.1 监控模块 5
2.1.1 传统监控中的监控实现 5
2.1.2 智能监控中的监控实现与视频编解码技术 6
2.2 图像处理模块 8
2.2.1 人脸识别实现 8
2.2.2 图片匹配算法 9
2.3 语音合成模块 10
2.4 非功能性模块 12
第3章 系统分析 13
3.1 系统概述 13
3.2 业务流程分析 13
3.3 功能需求分析 15
3.4 非功能需求分析 16
3.5 可行性分析 16
第4章 系统总体设计 18
4.1 系统结构设计 18
4.2 系统功能设计 18
4.3 系统界面设计 19
4.3.1 主界面设计 19
4.3.2 智能监控界面设计 20
4.3.3 Web页面设计 21
4.3.4 H5页面设计 22
4.3.5 Android界面设计 22
4.4 数据库设计 23
4.5 功能模块设计 23
4.5.1智能监控模块设计 23
4.5.2 网络监控模块设计 24
4.5.3 传统监控模块设计 24
4.5.4 图片匹配模块设计 24
第5章 系统测试和运行结果 25
5.1 本地系统测试 25
5.2 云端界面测试 27
5.3 本地系统测试结果 28
5.3.1 智能监控系统运行结果 28
5.3.2 网络监控模块运行结果 29
5.4 云端界面测试结果 29
5.4.1 网页运行结果 29
5.4.2 安卓端运行结果 30
第6章 总结与展望 31
6.1 研究总结 31
6.2 未来展望 31
参考文献 32
致 谢 34
第1章 绪论
1.1 研究背景与意义
随着基础设施建设的不断提高,我国监控设施遍布大街小巷,无论是政府大楼还是院落乡野,“天网恢恢,疏而不漏”。但是在传统的监控系统中,做到的多是纯人工的监控模式。对于突发情况往往容易疏忽,因而难以及时发现,也难以及时进语音化的智能报警。
传统的监控系统需要人工对视频进行多次的存储回放操作,这对于服务器的性能提出了非常高的要求,不利于监控工作的进行。但监控“天网”密布,设备猝然更换也会造成不小的资源浪费。因此在不对现有的旧监控设备改造的基础上,对监控系统加以改进,智能监控系统应运而生。
请支付后下载全文,论文总字数:26127字