Dissertation > Excellent graduate degree dissertation topics show

Research and Implementation of Android Face Detection System

Author: GuoChangYu
Tutor: LiGaoPeng
School: Harbin Institute of Technology
Course: Electronics and Communication Engineering
Keywords: Android Platform Face Detection Adaboost OpenCv
CLC: TP391.41
Type: Master's thesis
Year: 2013
Downloads: 7
Quote: 0
Read: Download Dissertation

Abstract


As the mobile Internet comes, Android system develops faster. and faster Currently,Android is already the world’s most popular mobile platforms. Mobile photography isour most popular demand, we hope to shine out of photos better, digital cameras canachieve automatic face focus, thereby improving the photographic effect. Generalnon-intelligent platform can not be achieved or improved face detection rate. But on theAndroid platform face detection rate can be changed through transplanting ClassLibrary. Currently, the face detection of the system is based on detection of human eyes,so the detection rate is not high. Therefore, this paper studies how to improve the humanface detection rate on the Android platform, This technology can be widely used toenhance the photographic effect.This paper introduces the research status of face detection, then select adaboostface algorithms based on statistical features.Main idea of adaboost face detectionalgorithm is for the rich facial expression samples, training weak classifiers, and thenforming strong classifier by weak classifiers, strong classifier is the best weakclassifiers through layers of screening. The best week classifiers that have lower erroroccupy larger proportion,while these that have higher error occupy smallerproportion.But only the strong classifier is not enough, we need make strong classifierform a cascade classifier. This detection rate is very high, because a picture which goesthrough the final adoption of the multi-layer authentication can only say that it is peopleface sample.Part job of the system is to implement strong classifier on the PC, another part isto achieve face detection on the phone. The first important thing is to choose a good partof the positive samples and negative samples for the training, positive and negativesamples should contains a variety of facial expressions and complex situations. Throughtraining, we can get cascade classifier XML file, then realize Adaboost algorithm. Thesecond part is that the Java code transfer pictures and trained cascade classifier XMLfile to a local method through JNI layer. Native method return face position by thepicture and cascade classifier XML file, which is partly realized in OpenCv inside.when the results are returned, Java layer code draw face window by calling the SDK.Android platform is Linux-embedded operating system, turn up package Lib library,application-layer framework, the application layer. Android developers can not directlyuse the C++language, while local codes and OpenCv code are all written by C++language, so cross-compiling NDK is used. NDK is a cross-compiler, we use the NDKto compile native code and form a dynamic link library, you can use it in the embeddeddevice such as phone.Face detection algorithm of Android system is based on thegeometric characteristics of the detection methods, it mainly detects eyes, if eyes are ata shelter, detection effect is decreased significantly. Detection rate of this system can reach90%, Face detection rates have made great progress compared with Androidsystem.

Related Dissertations

  1. Research on the Classification Based on the Reconstruction of Solder Joint,TP391.41
  2. Tongue Feature Extraction and Research of Fusion Classification,TP391.41
  3. The Fatigue State Recognition of the Driver Based on Eye Detection,TP391.41
  4. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  5. The Research on Paper Currency Classification Method Based on Harr-Like Feature and Minimal Ball Including Samples,TP391.41
  6. Research on Predicting Intrinsic Disorder Protein Structure Based on Supervision Manifold Learning Algorithm,Q51
  7. Face Recognition Method Based on DE,TP391.41
  8. Research on Face Detection Based on Skin Color Segmentation and AdaBoost Algorithm,TP391.41
  9. The Design and Implementation of Face Detection Algorithm Based on FPGA Chip,TP391.41
  10. Research and Implementation of Non-rigid Medical Image Registration Method Based on Improved Optical Flow Model,TP391.41
  11. Face Detection and Implementation on DSP,TP391.41
  12. Design and Realization of Embedded Face Detection and Tracking System,TP274
  13. Research & Implementation of AdaBoost-Based Vehicle Recognition System,TP391.41
  14. The Reserch of Rapid Face Detection and Real-time Tracking Algorithms Based on Video Streaming,TP391.41
  15. The Study of Fatigue Detection of Drivers Based on the Infrared Conditions,TP391.41
  16. Research on Face Detection System,TP391.41
  17. Research on Face Tracking Technologies Based on Movement Prediction,TP391.41
  18. Research & Implementation of Face Detection Algorithm Based on Adaboost,TP391.41
  19. Based on the Deep Web of Books Information Integration and Inquires the System,TP311.52
  20. Face Detection and Optical Flow Tracking in Video Sequences,TP391.41
  21. Face Detection Approaches Research Based on AdaBoost,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile