Download link: https://pan.baidu.com/s/1XjmD2mx9kP841NsBLX938Q Password: kfgr
- Publisher: Electronics Industry Press; 1st Edition (August 1, 2017)
- Paperback: 234 pages
- Language: Simplified Chinese
- Format: 16
- ISBN: 7121321386, 9787121321382
- Barcode: 9787121321382
- Product size: 23.8 x 17 x 1.6 cm
- Item weight: 381 g
- Brand: Electronic Industry Press
- ASIN: B074128LD9
Editor's Choice
"Behind the Face Brush: Face Detection, Face Recognition and Face Retrieval" Editor's recommendation: Face recognition is a popular research and development direction today, and has a wide range of applications in security, finance, tourism and other fields. "Behind Face Brushing: Face Detection, Face Recognition and Face Retrieval" comprehensively and systematically introduces the technology behind "face brushing", including algorithm principles and implementation techniques related to face detection, face recognition, and face retrieval.
About the Author
Zhang Zhongsheng, male, Ph.D., professor, master tutor, leader of the Big Data Research Center and Big Data Team of Henan University. His research fields are big data analysis, deep learning, data mining, database, data flow (real-time data analysis).
Graduated from INRIA, France (French National Institute of Information and Automation) with a Ph.D., and won the honor of outstanding doctoral dissertation. From August 2010 to March 2011, at the University of California, Los Angeles (UCLA), Department of Computer Science, under the tutelage of Professor Carlo Zaniolo, a famous database expert, engaged in cooperative research in the field of data mining. 2012-2013, Norwegian University of Science and Technology, ERCIM/Marie-Curie Fellow.
content
Chapter 1 Overview of Face Detection, Face Recognition and Face Retrieval 1
1.1 Application Scenarios of Face Detection, Face Recognition and Face Retrieval 2
1.1.1 Current Applications 3
1.1.2 Future Applications 5
1.2 Face Detection, People Commonly used datasets for face recognition and face retrieval 5
1.2.1LFW dataset 5
1.2.2FDDB dataset 6 1.2.3Wanwan1
dataset 7 1.2.4Wanwan2
dataset 8
1.3 Introduction, installation and use of OpenCV Image Processing Fundamentals 16 2.1 Fundamental Concepts of Digital Image Processing 16 2.1.1 Pixels 17 2.1.2 Resolution 17 2.1.3 Hue, Brightness and Saturation of Images 19 2.1.4 Contrast of Images 22 2.1.5 Textures of Images 23 2.2 Color space 26 2.2.1RGB color space 26 2.2.2HSV color space 27 2.2.3YUV color space 27 2.2.4 Color space conversion 28 2.3 Basic operations of digital image processing 32 2.3.1 Image reading 32 2.3.2 Image Display 34 2.3.3 Image Modification 35 2.3.4 Image Storage 36 2.3.5 Obtaining Basic Information of Image 37 2.4 Image Type and Conversion 38 2.4.1 Image Type 39 2.4.2 Image Type Conversion 39
2.5 Image transformation processing 48
2.5.1 Image translation 48
2.5.2 Image rotation 51
2.5.3 Image scaling 52
2.5.4 Image cropping 55
2.5.5 Image flip 58
2.6 Image noise and filtering 60
2.6. 1 Common Noise Models 60
2.6.2 Classical Denoising Algorithms 64
Chapter 3 Face Detection Practice 67
3.1 DPM Face Detection Algorithm 67
3.1.1 Use of DPM Face Detection Algorithm 68
3.1.2 Principles of DPM Face Detection Algorithm 69
3.1.3 The detection results of the DPM face detection algorithm 73 3.2 The
LAEO face detection algorithm 74
3.2.1 The use of the
LAEO face detection algorithm 74 3.2.2 The principle of the
LAEO face detection algorithm 75 3.2.3 The detection results of the LAEO face detection algorithm 77
3.3 Viola & Jones Face Detection Algorithm 79
3.3.1 Use of Viola & Jones Face Detection Algorithm 79 3.3.2
Principle of Viola & Jones Face Detection Algorithm 79 3.3.3
Detection Results of Viola & Jones Face Detection Algorithm 82
References 83
Chapter 4 Face Detection Based on Deep Learning Algorithm 84
4.1CNN Facial Point Detection Face Detection Algorithm 84
4.1.1 Use of CNN Facial Point Detection Face Detection Algorithm 85
4.1.2 Principle of CNN Facial Point Detection face detection algorithm 85
4.1.3 Detection results of CNN Facial Point Detection face detection algorithm 86
4.2DDFD face detection algorithm 87
4.2.1 The use of DDFD face detection algorithm 87
4.2.2DDFD face detection Principle of the algorithm 88
4.2.3 Detection results of DDFD face detection algorithm 89
4.3 Face detection algorithm fusion 90
References 92
Chapter 5 Face detection based on Fast R-CNN 94
5.1 Fast R-CNN introduction 94
5.2 Fast R-CNN Features and structure 95
5.3 Fast R-CNN use 96
5.4 Data set preprocessing 97
5.5 EdgeBoxes use 98
5.6 Use EdgeBoxes to extract object proposals99
5.7 Fast R-CNN based face detection network model training and testing 100
5.7.1 Training Stage 100
5.7.2 Testing Stage 106
5.7.3 Evaluation Stage 108
5.7.4 Optimization Stage 111
References 112
Chapter 6 Face Recognition Practice 113
6.1 DeepID Algorithm 114
6.1.1 Principles of
DeepID Algorithm 114 6.1.2 Process of DeepID Algorithm 116
6.1 .3 Results of the DeepID Algorithm 126
6.2VGG Face Descriptor Algorithm 128
6.2.1 Principle of VGG Face Descriptor Algorithm 128 6.2.2 Implementation of VGG Face
Descriptor Algorithm 129
6.2.3 Results of VGG Face Descriptor Algorithm 131
6.3 Three Face Recognition Algorithms in OpenCV 132
6.3.1Eigenfaces132
6.3.2Fisherfaces140
6.3.3 Local Binary Patterns Histograms148
6.4 Comparative Analysis of Face Recognition Algorithms 152
6.5 Summary 153
References 155
Chapter 7 Face Retrieval Practice 157
7.1 Introduction to Face Retrieval 157
7.2 Methods for Calculating Face Similarity 158
7.2.1 Euclidean Distance 159
7.2.2 Cosine similarity 159
7.3 Query processing algorithm 161
7.4 Criteria for evaluating face retrieval results 161
7.5 PHash algorithm 161
7.5.1 The use of PHash algorithm 162
7.5.2 PHash algorithm principle 162
7.5.3 PHash algorithm implementation 162
7.5.4 PHash algorithm experiment Data, Experimental Results and Analysis 164
7.6DHash Algorithm 168
7.6.1The Use of DHash Algorithm 168
7.6.2DHash Algorithm Principle 168
7.6.3DHash Algorithm Implementation 169
7.6.4 The experimental data, experimental results and analysis of the
DHash algorithm 170 7.7 The PCA algorithm 173
7.7.1 The use of the PCA algorithm 173 7.7.2 The
principle of the PCA algorithm 174
7.7.3 The realization of the PCA algorithm 175
7.7.4 The experimental data, experimental results and analysis of the PCA algorithm 177
7.8 BoF features 181
7.8.1 The use of the BoF
-SIFT algorithm 182 7.8.2 The principle of the
BoF-SIFT algorithm 182 7.8.3 The implementation of the BoF-SIFT algorithm 182
7.8.4 The experimental data, experimental results and analysis of the BoF-SIFT algorithm 188
7.9 For fast image retrieval KD-Tree Index 190
7.9.1 FLANN Algorithm Use 191
7.9.2 KD-Tree Creation and Query Processing 191
7.9.3 KD-Tree Algorithm Implementation in FLANN 192
7.9.4 Experimental Data, Experimental Results and Analysis of FLANN Algorithm 194
7.10 Gabor Algorithm 195
7.10.1 The use of the
Gabor algorithm 196 7.10.2 The principle of the Gabor algorithm 196
7.10.3 The implementation of the Gabor algorithm 199
7.10.4 The experimental data, experimental results and analysis of the Gabor algorithm 204 7.11 The
HOG algorithm 208
7.11.1 The use of the
HOG algorithm 209 7.11.2 The principle of the HOG algorithm 209
7.11.3 HOG Algorithm Implementation 210
7.11.4 Experimental Data, Experimental Results and Analysis of HOG Algorithm 212
7.12 Deep Learning Features 215
7.12.1 Use of Deep Learning Algorithms 215
7.12.2 Principles of Deep Learning Algorithms 215
7.12.3 Implementation of Deep Learning Algorithms 216
7.12.4 Experimental Data, Results and Analysis of Deep Learning Algorithms 216
References 220
No. 8 Chapter Face Detection Commercial Software and Application Examples 222
8.1 VeriLook of Face
Detection Commercial Software 222 8.2 Face++ of Face Detection Commercial Software 226
8.3 Comparative Analysis of Various Face Detection Algorithms 229
8.4 Face Detection and Tracking in Video 231
References 234
Preamble
我们正处于“刷脸”的时代，越来越多的“刷脸”应用开始出现。例如，北京西站的刷脸检票、厦门景点的刷脸验票、余额宝的刷脸认证等。初学者如果想进行人脸识别相关的研究和开发，那么他们应该阅读什么书籍呢？
“刷脸”背后的技术，不仅仅是人脸识别，亦需要人脸检测和人脸检索等技术提供支撑。目前，市场上有少部分人脸识别的书籍，而专门讲解人脸检测和人脸检索技术的书籍则更少。近年来，笔者及其团队在从事人脸检测、人脸识别、人脸检索相关的研究时，查阅了很多国内外的参考资料，到目前为止，尚未见到一本能够全面涵盖“刷脸”应用所涉及的人脸检测、人脸识别和人脸检索相关技术且有实战参考价值的书籍。其中的一个主要原因可能是刷脸技术的商业价值高。
本书按照“刷脸”应用开发时所需技术的先后顺序，通过原理、例子、实战的方式，分别讲解了“刷脸”应用需要掌握的三大技术：人脸检测、人脸识别和人脸检索。更为重要的是，本书高度注重实战应用，每一个算法都通过具体程序讲解算法的使用、实验设计，以及实验结果。读者不但能够了解每个算法的原理，而且能够掌握应用开发的实战技能。
本书的目标是作为通用、普及性强、可操作性强的人脸识别的书籍，方便研究人员、工程师、研究生、计算机专业的高年级本科生，快速上手并全面、深入理解，扎实掌握“刷脸”应用相关的理论和算法，帮助读者快速入门，理解“刷脸”应用背后的核心技术与算法，并切实掌握“刷脸”应用开发所需的实战技术。
本书主编为张重生，副主编为王弯弯、王朋友、赵冬冬。于珂珂、彭国雯、裴宸平等研究生对本书的编写、实验部分的验证提供了一定的帮助，在此致谢。
笔者自知才疏学浅，仅略知人脸检测、人脸识别、人脸检索之皮毛。书中错谬之处在所难免，如蒙读者不吝告知，将不胜感激。
张重生
2017年4月