Phân tích đa phân giải xây dựng thuật toán giám định ảnh cho ảnh copy move

168 6 0
Phân tích đa phân giải xây dựng thuật toán giám định ảnh cho ảnh copy move

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

Thông tin tài liệu

ĐẠI HỌC QUỐC GIA TP HỒ CHÍ MINH TRƯỜNG ĐẠI HỌC BÁCH KHOA HUỲNH KHẢ TÚ PHÂN TÍCH ĐA PHÂN GIẢI XÂY DỰNG THUẬT TOÁN GIÁM ĐỊNH ẢNH CHO ẢNH COPY-MOVE LUẬN ÁN TIẾN SĨ KỸ THUẬT TP HỒ CHÍ MINH NĂM 2018 ĐẠI HỌC QUỐC GIA TP HCM TRƯỜNG ĐẠI HỌC BÁCH KHOA PHÂN TÍCH ĐA PHÂN GIẢI XÂY DỰNG THUẬT TOÁN GIÁM ĐỊNH ẢNH CHO ẢNH COPY-MOVE Chuyên ngành: Kỹ thuật Điện Tử Mã số chuyên ngành: 62520203 Phản biện độc lập 1: PGS.TS.Lê Thị Lan Phản biện độc lập 2: PGS.TS.Dương Anh Đức Phản biện 1: GS.TSKH.Nguyễn Ngọc San Phản biện 2: GS.TS Vũ Đình Thành Phản biện 3: PGS.TS.Nguyễn Hoàng Hải NGƯỜI HƯỚNG DẪN KHOA HỌC GS.TS Lê Tiến Thường TS Hà Việt Uyên Synh LỜI CAM ĐOAN Tác giả xin cam đoan cơng trình nghiên cứu thân tác giả Các kết nghiên cứu kết luận Luận án trung thực, không chép từ nguồn hình thức Việc tham khảo nguồn tài liệu thực trích dẫn ghi nguồn tài liệu tham khảo theo yêu cầu Tác giả Luận án Huỳnh Khả Tú TĨM TẮT LUẬN ÁN Luận án trình bày nội dung thực nghiên cứu, từ khảo sát giám định ảnh Copy-Move đến đề xuất thuật toán giám định ảnh khả thi, hướng tới ứng dụng phân tích đa phân giải để cân độ xác thời gian xử lý, đáp ứng yêu cầu mục tiêu đặt Với nghiên cứu khảo sát từ kỹ thuật liên quan lĩnh vực giám định ảnh Copy-Move, Luận án xây dựng sơ đồ tổng quát để giải toán Xác định vùng giống nhau, sở thao tác Copy-Move, sử dụng dựa phân tích đặc tính khối ảnh nhỏ tạo nên vùng chép Trên sở đó, Luận án đề xuất giải thuật cho thuật toán giám định ảnh CopyMove cách tìm vùng giống sau tách dựa phân tích histogram Giải thuật tách khỏi đối tượng để tìm vùng chép nhằm giảm dư thừa tính tốn vùng ảnh có tương đối đồng đối tượng đơn giản Với ảnh bất kỳ, cách trích đặc tính dùng Zernike Moments cải tiến (Modified Zernike Moments-MZMs) nhằm cực đại hóa số pixels ánh xạ vào vòng tròn đơn vị đưa giải thuật đề xuất thứ giảm sai số hình học sai số số học tìm vùng giống Hai giải thuật ban đầu đưa từ nghiên cứu khảo sát, giúp cải thiện độ phức tạp (đối với giải thuật đầu) độ xác (đối với giải thuật thứ 2) cách xét trực tiếp ảnh, thơng qua đó, cho thấy cần thiết để phát triển đa phân giải vào việc xây dựng thuật toán giám định ảnh Copy-Move hiệu Từ nhận định trên, Luận án tiến hành nghiên cứu lý thuyết hai phương pháp phân tích đa phân giải bao gồm wavelets curvlets; đồng thời xây dựng thuật toán thứ để tìm wavelets ứng viên bật cho giám định ảnh Copy-Move Không thế, đánh giá so sánh kích thước khối ảnh, lọc cho wavelets, bậc Zernike Moments (ZMs) xác định từ thuật toán thứ Giải thuật thứ đề xuất cho thuật toán giám định ảnh Copy-Move cách phát vùng giống từ vectors đặc tính dùng MZMs thành phần xấp xỉ LL1 biến đổi wavelets rời rạc, cân thời gian xử lý độ xác tập ảnh đa dạng Giải thuật đồng thời xác định vùng gốc vùng chép dựa vào so sánh giá trị độ sắc vùng phát giống nhau, mà phương pháp trước chưa đề cập đến Để tận dụng dải thành phần tần số cao, Luận án xây dựng giải thuật thứ đề xuất cho thuật toán giám định ảnh Copy-Move cách so sánh đặc tính từ thành phần xấp xỉ tính tốn độ sắc thành phần tần số cao DWT Tính tốn độ sắc thành phần tần số cao với mục đích xác định vị trí vùng cắt dán nghi ngờ, kết hợp với vùng giống tìm từ so sánh đặc tính dùng Phương pháp khác biệt chuyển động (Run Difference Method-RDM) thành phần xấp xỉ giới hạn vùng chép cải tiến độ xác Với đặc tiểm vị trí có cắt dán độ sắc cao khác biệt với cạnh vốn có ảnh, Luận án đưa đề xuất cho thuật toán xác định giả mạo CopyMove Splicing từ việc tính tốn độ sắc thành phần tần số cao, kết hợp khôi phục vùng giả mạo Copy-Move từ thành phần xấp xỉ biến đổi DWT Với nội dung nghiên cứu này, giải thuật trước tiên xác định ảnh có giả mạo hay khơng, trường hợp có giả mạo thao tác giả mạo (bao gồm Copy-Move Splicing) đối tượng giả mạo tương ứng xác định Việc xác định thao tác Splicing điểm liên quan xác định thao tác, đề tài không nghiên cứu giải thuật cho ảnh Splicing Một hệ thống nhúng dùng Raspberry Pi3B triển khai thực giải thuật khẳng định hiệu khả ứng dụng giải thuật Các giải thuật triển khai thuật toán giám định ứng dụng đa phân giải kiểm chứng ảnh Copy-Move trích từ tập ảnh benchmark [1], từ ảnh tự nhiên với thao tác Photoshop, trích từ dbforgery [2] có Splicing (đối với giải thuật 6) thơng qua mơ Matlab 2013a, Windows Ultimate 64-bit, CPU Intel Core i5 @ 1.8GHz and 4GB RAM Kết giải thuật đánh giá so sánh độ xác, recall F với phương pháp liên quan mức pixels Luận án tập hợp kết 13 cơng trình nghiên cứu, 05 cơng trình đăng Tạp chí uy tín (02 Tạp chí indexed Thomson Reuters, Web of Science Scopus, 02 Tạp chí thuộc danh mục tính điểm Hội đồng học hàm, 01 Tạp chí quốc tế có phản biện), 01 chương sách 07 báo cáo khoa học trình bày Hội nghị khoa học quốc tế ABSTRACT The dissertation presents contents of the research process, from doing survey on Copy-Move forgery detection (CMFD) to proposing the suitable image forensics algorithms, oriented to applying multi-resolution analysis to balance the accuracy and processing time, which meets the given requirements and objectives With research studies on the field of Copy-Move image forensics, the dissertation builds a general diagram to solve the problem Identifying the similar areas, which is the basis of the Copy-Move operation, is used based on the analysis of the characteristics of the small blocks that constitute the replication area On that basis, the dissertation proposes the first algorithm for the CMFD algorithm by finding the copied regions after eliminating the background using histogram analysis The fact that the algorithm separates the background from the foreground before finding the replication region reduces the computing redundant of backgournd in case of relatively uniform background images and simple objects In an arbitrary image, feature extraction using modified Zernike Moments (MZMs) to maximize the number of pixels mapped to the unit circle, which is suggested in the second proposed algorithm, reduces the geometric and numerical error when searching the duplicated regions The first two algorithms suggested from investigational studies not only improve the complexity (for the first algorithm) and accuracy (for the second algorithm) by processing directly in the tested image but also show the necessary to develop a multiresolution analysis in building the effective CMFD algorithms From the above suggestion, the dissertation studies on theory of multiresolution analysis of wavelets and curvelets; and also build the third algorithm from which wavelets is proved to be more suitable candidate in image forensics In addition, comparisons of image block size, wavelets filters, order of Zernike Moments (ZMs) are also discussed in this algorithm The fourth algorithm proposes a CMFD solution by detecting the similar regions from the characteristic vectors using MZMs at the LL1 approximation of Discrete Wavelet Transforms, which balances between processing time and accuracy in the more diverse set of images The algorithm also identifies the original and copied regions based on the comparison of the value of the sharpness in the detected areas, which have not been previously mentioned In order to take advantage of high frequency components, the dissertation builds the fifth algorithm which proposes for the Copy-Move image detection by comparing properties from the approximation component and estimating the sharpness at high frequency components of DWT Sharpness estimation at high frequency components to identify the locations of suspected copied-pasted regions, coupled with similar regions detected from the comparison of characteristics using the Run Difference Method (RDM) at the approximation component, can limit the copied areas and improve the accuracy With the fact that the sharpness at the pasted position is very clear and different from the edges of the image, the dissertation proposes the sixth algorithm for the Copy-Move or Splicing counterfeit detection by sharpness estimation at high frequency components, combined with Copy-Move faked region reconstruction from the approximate component of the DWT transform In this research, the algorithm firstly determines whether the image is tampered; and in the case of tampering, the manipulation of tampering (including Copy-Move and Splicing) and corresponding fake objects will be defined Determining the Splicing manipulation is only one of the points involved in determining the manipulations, and the dissertation does not study the algorithm for the Spliced images An embedded system using Raspberry Pi3B to implement the proposed algorithm has confirmed the efficiency and applicability of the algorithm The algorithms deploying multi-resolution applications are tested on the CopyMove images from a set of benchmark images [1], from natural images with Photoshop manipulations, or from the dbforgery [2] in case of Splicing (for algorithm 6) by Matlab 2013a, on Windows Ultimate 64-bit, Intel Core i5 processor @ 1.8GHz and 4GB RAM The results of proposed algorithms are compared in accuracy, recall and F with related methods at pixel level The dissertation is a collection of 13 papers, of which 05 papers are published in prestigious journals, 01 book chapter and 07 scientific papers presented at the International Conferences LỜI CÁM ƠN “Chưa biết tri ân chưa biết sống”- câu nói tình cờ nghe, trở thành châm ngôn sống Trong sống này, người có khả năng, ý chí phấn đấu, tâm, để đạt thành tựu riêng; nhiên điều quan trọng, yếu tố khơng thể thiếu để người tiếp cận đến mục tiêu thật trọn vẹn, dẫn dắt định hướng mơi trường có nhiều điều kiện để nắm bắt nhiều hội tích lũy kiến thức lĩnh vực hướng tới Với thành người đạt khác dù mức độ nào, “tri ân’ điều không nhớ tới Mục tiêu tơi khơng cao xa, tơi đạt không to lớn Nhưng Luận án ấp ủ mong đợi Để đặt bút viết dịng này, điều trước tiên tơi muốn nói, kính gửi đến GS.TS Lê Tiến Thường lời tri ân chân thành Thầy dẫn dắt, định hướng, động viên tạo điều kiện tốt để tơi hồn thành nghiên cứu Với kết mà tơi đạt được chia sẻ, khích lệ Thầy tư vấn để phát triển ý tưởng Tôi may mắn học Thầy từ học chuyên ngành tiếp tục nghiên cứu hướng dẫn đầy trách nhiệm nhiệt huyết Thầy Tôi kính gửi đến TS Hà Việt Uyên Synh với hỗ trợ, chia sẻ ý kiến có ý nghĩa cho q trình nghiên cứu tơi Tơi chân thành cảm ơn Thầy/Cô Bộ Môn Viễn Thông có ý kiến đóng góp chun mơn chun đề để tơi cải tiến hồn thiện nghiên cứu; Cám ơn Thầy/Cô/Anh/Chị Khoa Điện-Điện Tử, Phịng Đào tạo Sau đại học ln tạo điều kiện thuận lợi cho tôi, tư vấn hữu ích q trình thực nghiên cứu Tôi trân trọng cám ơn Trường Đại học Bách Khoa, Đề án 911 nơi công tác cho tơi hội để thực hồn tất Luận án Cám ơn người bạn, đồng nghiệp, người ln động viên tơi hồn thành Luận án Kính gửi đến Ba, Mẹ em trai cho tất tơi có đến ngày hôm nghiệp sống Tp.HCM, ngày 06/8/2018 MỤC LỤC MỞ ĐẦU 18  Sự cần thiết nghiên cứu 19  Đặt vấn đề 22  Hướng giải vấn đề 22  Những đóng góp Luận án 24  Cấu trúc Luận án 28 CHƯƠNG 1: TỔNG QUAN NGHIÊN CỨU LIÊN QUAN VÀ ĐỀ XUẤT GIẢI THUẬT NGHIÊN CỨU BAN ĐẦU 30 1.1 Tổng quan nghiên cứu liên quan 31 1.1.1 Khảo sát phân tích giải thuật liên quan đến đề tài 31 1.1.2 Những vấn đề tồn từ giải thuật khảo sát phân tích 46 1.1.3 Kết luận tình hình nghiên cứu liên quan định hướng xây dựng nghiên cứu ban đầu 47 1.1.4 Các thông số đánh giá sử dụng giải thuật giám định ảnh Copy-Move 48 1.2 Các giải thuật nghiên cứu ban đầu 48 1.2.1 Thuật toán giám định ảnh Copy-Move cách tìm vùng giống sau tách dựa phân tích histogram 48 1.2.2 Thuật toán giám định ảnh Copy-Move dùng Zernike Moments (ZMs) với sai số hình học sai số số học cải tiến 60 1.3 Đánh giá so sánh 02 giải thuật đề xuất giải thuật liên quan 72 1.4 Kết luận Chương 73 CHƯƠNG 2: PHÂN TÍCH ĐA PHÂN GIẢI VÀ CÁC THUẬT TOÁN ỨNG DỤNG ĐA PHÂN GIẢI TRONG GIÁM ĐỊNH ẢNH COPY-MOVE 75 2.1 Phân tích đa phân giải wavelets, curvelets 77 2.1.1 Wavelets 77 2.1.2 Curvelets 81 2.1.3 Nhận xét 84 2.2 Thuật toán giám định ảnh Copy-Move kết hợp trích đặc tính dùng Zernike moments đa phân giải 85 2.2.1 Mơ tả tốn 85 2.2.2 Giải thuật giám định ảnh Copy-Move kết hợp trích đặc tính dùng Zernike moments đa phân giải 85 2.2.3 Kết thử nghiệm cho giải thuật giám định ảnh Copy-Move kết hợp trích đặc tính dùng Zernike moments đa phân giải 87 2.2.4 So sánh wavelets curvelets với vai trò đa phân giải giám định ảnh Copy-Move 92 2.2.5 Kết luận giải thuật 96 2.3 Các thuật toán đề xuất dùng DWT MZMs 97 2.3.1 Biến đổi Wavelets rời rạc giảm độ phức tạp tính tốn 97 2.3.2 Thuật toán giám định ảnh Copy-Move cách phát vùng giống từ thành phần xấp xỉ LL1 DWT phân biệt vùng chép vùng gốc 100 2.3.3 Thuật toán giám định ảnh Copy-Move cách so sánh đặc tính LL1 tính tốn độ sắc HH1 DWT 113 2.4 Đánh giá so sánh 03 giải thuật đề xuất giải thuật liên quan 121 2.5 Kết luận Chương 123 CHƯƠNG 3: THUẬT TOÁN XÁC ĐỊNH THAO TÁC GIẢ MẠO VÀ KHÔI PHỤC VÙNG GIẢ MẠO DO COPY-MOVE 124 3.1 Mô tả toán 126 3.2 Thuật toán xác định thao tác giả mạo copy-move Splicing từ thành phần tần số cao, kết hợp khôi phục vùng giả mạo copy-move từ thành phần xấp xỉ 126 vùng giống thật làm giảm độ xác giải thuật, giải thuật 1, ưu tiên giảm độ phức tạp tính tốn Giải thuật chí giảm độ xác mạnh Hình i.1 (c) đối tượng tương đối giống việc tách cắt thông tin đối tượng trước tìm vùng giống Khi so sánh thời gian phát giả mạo Copy-Move từ kết Bảng i.2 Hình i.3, nhận thấy với 06 giải thuật đề xuất, giải thuật 1, cho thời gian xử lý nhanh giải thuật cịn lại sử dụng tách trước tìm vùng giống giảm bớt thơng tin, dẫn đến giảm độ phức tạp tính toán thời gian xử lý Giải thuật thứ (GT2) có thời gian xử lý cao hợp lý việc sử dụng MZMs trích đặc tính để tìm vùng giống gia tăng độ phức tạp tính tốn thời gian xử lý Đối với nhóm giải thuật sử dụng DWT mức với vai trò đa phân giải (GT 4,5,6), thời gian cải thiện rõ; giải thuật (GT6) tính hai giai đoạn bao gồm phát giả mạo xác định thao tác có thời gian cao giá trị thể hình vẽ so với nhóm Giá trị thể hình tính giai đoạn phát giả mạo Hình i.4 Kết so sánh kết hợp thời gian xử lý trung bình độ xác trung bình giải thuật thực hình biểu diễn Hình i.1 Xét đến cân thời gian phát ảnh có giả mạo độ xác giải thuật Luận án đề xuất đạt mức độ định Tuy nhiên, việc sử dụng giải thuật theo mục đích khác chọn lựa khác Đối với việc 152 xử lý tập ảnh lớn, mục tiêu cần phân loại ảnh có giả mạo hay khơng, ưu tiên thời gian xử lý, giải thuật 1, (nếu xét đến giai đoạn phát GT6) quan tâm Đối với việc u cầu độ xác GT2 ý Trong trường hợp cân thời gian độ xác, giải thuật 4,5 chiếm ưu thế, giải thuật (GT5) xem tối ưu (như Hình i.4) Bên cạnh việc xác định giả mạo, giải thuật (GT4) phân biệt vùng chép vùng gốc, điểm nghiên cứu công bố liên quan đến giám định ảnh Copy-Move Việc triển khai hệ thống nhúng ban đầu dùng Raspberry Pi3B chứng minh tính ứng dụng giải thuật (GT6), triển khai để làm phần thực nghiệm cho sinh viên bước đầu nghiên cứu ứng dụng giám định ảnh, tạo tảng cho nghiên cứu sâu  Các công bố khoa học Luận án Luận án thực q trình liên tục với cố gắng khơng ngừng, kết đạt với mục tiêu Luận án đưa Mỗi giai đoạn nghiên cứu tập trung vào giải thuật đề xuất cụ thể, bên cạnh khơng ngừng cập nhật cơng bố lĩnh vực nghiên cứu Các giải thuật đề xuất tương ứng với thuật tốn có mối quan hệ nhau, thể phát triển cải tiến từ kết thu trình để tạo thành quy trình nghiên cứu có hệ thống hiệu Kết trình bày Luận án tổng hợp 13 cơng trình nghiên cứu, có 05 cơng trình đăng Tạp chí (02 tạp chí quốc tế indexed Thomson Reuters (ESCI) Scopus, 01 tạp chí quốc tế có phản biện, 02 tạp chí thuộc danh mục tính điểm học hàm), 01 Chương sách 07 công bố khoa học Hội nghị khoa học quốc tế uy tín Bên cạnh đó, số nội dung Luận án đóng góp vào 03 đề tài nghiên cứu khoa học cấp sở (đã nghiệm thu) 01 đề tài cấp Đại học Quốc gia Loại B (đã nghiệm thu) trình thực Luận án Các nghiên cứu công bố Tạp chí Khoa học: Thuong Le-Tien, Tu Huynh-Kha, An Tran-Hong, Long Pham-Cong-Hoan, Nilanjan Dey, Marie Luong, “Combined Zernike Moment and Multiscale Analysis for Tamper Detection in Digital Images”, Informatica (An International Journal of Computing and Informatics), 153 Vol.41, No.1, March 2017, ISSN: 0350-5596, indexed by Thomson Reuters (ESCI) and Scopus Tu Huynh-Kha, Thuong Le-Tien, Synh Ha-Viet-Uyen, Khoa Huynh-Van, Marie Luong, “A Robust Algorithm of Forgery Detection in Copy-Move and Spliced Images”, (IJACSA) International Journal of Advanced Computer Science and Applications, Vol 7, No 3, 2016, NY, USA, ISSN 2158-107X, indexed by Thomson Reuters (ESCI) and Web of Science Tu Huynh-Kha, Thuong Le-Tien, Synh Ha-Viet-Uyen, Khoa Huynh-Van, “Improving the Computational Cost for Copied Region Detection in Forensic Images, Journal Issue on Information and Communications Technology, Da Nang University, Vol.2, No.1 , 2016, ISSN: 1859-1531 (tạp chí thuộc danh mục tạp chí tính điểm theo quy định Hội đồng chức danh giáo sư nhà nước) Thuong Le-Tien, Tan Huynh-Ngoc, Tu Huynh-Kha, "The total error limits in duplicated image by Modifying the paremeters of Zernike Moments computation", International Journal of Automation and Control Engineering, Dec.2015, ISSN: 2301-3702, Engineering and Technology Publishing, USA Tien Vo-Minh, Thuong Le-Tien, Tu Huynh-Kha, and Marie Luong, "An embedded system implemented the multiresolution-based algorithm for forensic image detection," Scientific Journal of Saigon University, Vol.26, No.3, 2017, ISSN: 1859-3208 (tạp chí thuộc danh mục tạp chí tính điểm theo quy định Hội đồng chức danh giáo sư nhà nước) Chương sách (Book Chapter): Thuong Le-Tien, Tan Huynh-Ngoc, Tu Huynh-Kha, Marie Luong, "Zernike Momentsbased approach for detecting duplicated image regions by a modified method to reduce geometrical and numerical error", book chapter, Springer 2015, http://link.springer.com/chapter/ 10.1007%2F978-3-319-21410-8_36, ISSN: 0302-9743, ISBN: 978-3-319-21409-2 (seleted paper from ICCSA, Canada 2015) Các nghiên cứu công bố Hội nghị Khoa học: Tu Huynh-Kha, Thuong Le-Tien, Synh Ha-Viet-Uyen, Khoa Huynh-Van, Son HuynhThanh, “Develop an algorithm for image forensics using feature comparison and sharpness estimation”, The IEEE 2017 International Conference on Recent Advances in Signal Processing, Telecommunications & Computing, Danang, Vietnam, Jan 2017, ISSN: 978-1-5090-2291-5 154 Thuong Le-Tien, Sinh Nguyen-Duc, Tu Huynh-Kha, “Blind Image Forgeries Detection using Detail Components from the Wavelet Transform”, the Seventh International Conference on the Applications Digital Information and Web Technologies, Keelung, Taiwan, Mar.29-31, 2016, ISBN: 978-1-61499-636-1 Thuong Le-Tien, Marie Luong, Tu Huynh-Kha, Long Pham-Cong-Hoan, An T.H, “Block Based Technique for Detecting Copy-Move Digital Image Forgeries: Wavelet Transform and Zernike Moments”, Proceedings of The Second International Conference on Electrical and Electronic Engineering, Telecommunication Engineering, and Mechatronics, Philippines, Feb.2016, ISBN: 978-1-941968-30-7 10 Tu Huynh-Kha, Thuong Le-Tien, Synh Ha-Viet-Uyen, Khoa Huynh-Van, Nga Ly-Tu, "Forgery Detection and Identification of the Original and Duplicate Region in Copy-Move Images", The International Symposium on Electrical and Electronics Engineering, HCMC, Vietnam, Oct 30, 2015 11 Tu Huynh-Kha, Thuong Le-Tien, Synh Ha-Viet-Uyen, Khoa Huynh-Van, "The Efficiency of Applying DWT and Feature Extraction into Copy-Move Images Detection", The IEEE 2015 International Conference on Advanced Technologies for Communications, HCMC, Vietnam, Oct 14-16 2015, ISSN: 2162-1020, ISBN: 978-1-4673-8372-1 12 Tu Huynh-Kha, Thuong Le-Tien, Khoa Huynh-Van, Sy Nguyen-Chi, "A survey on Image Forgery Detection Techniques", The 11th IEEE-RIVF International Conference on Computing and Communication Technologies", Can Tho, Vietnam, Jan 25-28 2015, ISBN: 978-1-4799-8043-7 13 Tu Huynh-Kha, Thuong Le-Tien, Ti Nguyen-Ti, Khoa Huynh-Van, "Copy-Move Forgery Detection Techniques: A survey and New Approach", The Third Asian Conference on Information Systems 2014, Nha Trang, Vietnam, Dec 1-3 2014, ISBN: 978-4-88686-089 Các đề tài nghiên cứu khoa học liên quan: Ngồi đóng góp dạng cơng bố khoa học, Luận án đóng góp vào đề tài nghiên cứu khoa học cấp Đại học Quốc gia cấp sở, Nghiên cứu sinh thành viên tham gia 01 đề tài nghiên cứu khoa học cấp Đại học Quốc Gia, Loại B, chủ nhiệm 02 đề tài nghiên cứu khoa học cấp sở, cụ thể sau  Đề tài nghiên cứu khoa học cấp Đại học Quốc Gia, Loại B “Phát triển thuật tốn phân tích đa phân giải vào giám định ảnh”; Mã số đề tài: B201520-02; Chủ nhiệm đề tài: GS.TS Lê Tiến Thường; Đề tài giao nhiệm vụ năm 2015 nghiệm thu vào tháng 12/2017 với kết Đạt 155  Đề tài nghiên cứu khoa học cấp sở (03 đề tài)  “Phát triển giải thuật giám định ảnh dựa phân tích độ sắc cạnh ảnh”; Mã số đề tài: T2016-01-IT; Chủ nhiệm đề tài: Huỳnh Khả Tú; Đề giao nhiệm vụ năm 2016 nghiệm thu vào tháng 11/2017 với kết Tốt  “Xây dựng thuật toán giám định ảnh cho ảnh giả mạo hình thức CopyMove”; Mã số đề tài: T2015-04-IT; Chủ nhiệm đề tài: Huỳnh Khả Tú; Đề tài giao nhiệm vụ năm 2015 nghiệm thu vào tháng 06/2016 với kết Tốt  “Phân tích ứng dụng moment Zernike phát ảnh giả mạo hình thức copy-rotate-move”; Mã số đề tài: TNCS-2014-ĐĐT-09; Chủ nhiệm đề tài: Huỳnh Khả Tú; Đề tài thuộc khuôn khổ dự án 911 cấp cho nghiên cứu sinh; Đề tài giao nhiệm vụ tháng 06/2014 nghiệm thu với vào tháng 12/2014  Hướng nghiên cứu Từ kết nghiên cứu, Luận án đề xuất hướng nghiên cứu phát triển sau:  Các giá trị ngưỡng sử dụng giới hạn khối có đặc tính tương tự giá trị chọn theo kiểm tra giá trị cố định, việc nghiên cứu phát triển ngưỡng động tùy theo thông tin bố cục ảnh xét đến  Nghiên cứu thuật tốn cho vùng ảnh chép có thay đổi tỉ lệ ảnh Copy-Move dùng phương pháp dựa vào điểm ảnh point-based nâng cao độ xác khơi phục vùng giả mạo  Ứng dụng Deep Learning vào giám định ảnh Copy-Move  Phát triển giải thuật hiệu cho ảnh Copy-Move có can thiệp nhiều thao tác chép, đánh giá so sánh bao quát dựa kết tổng hợp từ 2007 đến 2017 khảo sát công bố năm 2018 [77] 156 PHỤ LỤC PHỤ LỤC 1: Cấu hình chung để thực giải thuật Tất giải thuật đề xuất Luận án chạy Matlab2013 PC Intel(R) Core ™ i5-2400 CPU@3.10 GHz, RAM 4GB PHỤ LỤC 2: Giải thuật K-means Gồm bước sau: Khởi động ngẫu nhiên K tâm (center) K clusters Mỗi phần tử gán cho tâm gần với phần tử dựa vào khoảng cách Euclidean Cập nhật lại tâm K clusters, tâm giá trị trung bình (mean) phần tửtrong cluster Lặp lại bước 2,3 hội tụ Để tìm số clusters, thuật tốn K-means phải chạy lặp lại với số giá trị K so sánh kết Trong giải thuật đề xuất, giải thuậ t Kmeans sử dụng số liệu dùng để so sánh kết giá trị khác K khoảng cách trung bình điểm ảnh điểm cluster PHỤ LỤC 3: Các ảnh sử dụng thử nghiệm Luận án Các ảnh sử dụng Luận án trích từ tập ảnh benchmark (đối với ảnh Copy-Move) dbforgery (đối với ảnh Splicing) cho giải thuật 6, ảnh chụp tự nhiên, 03 ảnh từ MICC-F600 (dạng chép Copy-Move nhiều vị trí) PHỤ LỤC 3.1: Ảnh sử dụng chung cho giải thuật Các ảnh chọn để thực thử nghiệm cho giải thuật tính bao qt đa dạng Phần lớn giải thuật trước sử dụng đánh giá số ảnh phổ biến 12 ảnh tập ảnh bên (trích từ benchmark), 15 ảnh từ internet tự chụp 03 ảnh cuối từ MICC-F600 Đây ảnh tượng trưng, chọn phù hợp với yêu cầu toán đặt giải thuật 157 158 PHỤ LỤC 3.2: Ảnh sử dụng riêng cho giải thuật Giải thuật 1: Với mục tiêu xây dựng giải thuật tách tìm kiếm đối tượng chép ảnh có bố cục tương đối đồng nhất, ảnh chọn để thử nghiệm chủ yếu cho giải thuật chọn ảnh có bố cục đồng nhất, đói tượng chép đơn khơng có thao tác xoay Giải thuật 2: Giải thuật sử dụng MZMs, với đặc tính xoay hiệu ZMs, ảnh sử dụng giải thuật bao gồm ảnh tập ảnh chung, số ảnh riêng có đối tượng chép có khơng có thao tác xoay 159 Giải thuật 3: Giải thuật đưa so sánh wavelets curvelets Các ảnh sử dụng cho chương nhiều để đánh giá thuyết phục (đã giới thiệu riêng phần trình bày Mục 2.2.3 Trang 86) Giải thuật 4,5: Ngoài ảnh từ tập ảnh chung, giải thuật sử dụng thêm ảnh mức xám để chứng minh hiệu giải thuật Giải thuật 6: Các ảnh sử dụng cho giải thuật 6, ngồi tập ảnh chung, có thêm ảnh chép nhiều vị trí, ảnh Splicing số trích từ tập ảnh dbforgery_10 số ảnh tự thao tác ảnh Splicing chưa có nhiều tập ảnh chuẩn công bố sử dụng 160 TÀI LIỆU THAM KHẢO [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] V Christlein, C Riess, J Jordan, C Riess, and E Angelopoulou, "An evaluation of popular copy-move forgery detection approaches," IEEE Transactions on information forensics and security, vol 7, pp 1841-1854, 2012 S Battiato and G Messina, "Digital forgery estimation into DCT domain: a critical analysis," in Proceedings of the First ACM workshop on Multimedia in forensics, 2009, pp 37-42 G K Birajdar and V H Mankar, "Digital image forgery detection using passive techniques: A survey," Digital Investigation, vol 10, pp 226-245, 2013 T K Huynh, K V Huynh, T Le-Tien, and S C Nguyen, "A survey on image forgery detection techniques," in Computing & Communication TechnologiesResearch, Innovation, and Vision for the Future (RIVF), 2015 IEEE RIVF International Conference on, 2015, pp 71-76 W Luo, J Huang, and G Qiu, "Robust detection of region-duplication forgery in digital image," in Pattern Recognition, 2006 ICPR 2006 18th International Conference on, 2006, pp 746-749 X Kang and S Wei, "Identifying tampered regions using singular value decomposition in digital image forensics," in Computer Science and Software Engineering, 2008 International Conference on, 2008, pp 926-930 A Popescu and H Farid, "Exposing digital forgeries by detecting duplicated image region [Technical Report] 2004-515," Hanover, Department of Computer Science, Dartmouth College USA, p 32, 2004 H.-J Lin, C.-W Wang, and Y.-T Kao, "Fast copy-move forgery detection," WSEAS Transactions on Signal Processing, vol 5, pp 188-197, 2009 H C Nguyen and S Katzenbeisser, "Detection of copy-move forgery in digital images using radon transformation and phase correlation," in Intelligent Information Hiding and Multimedia Signal Processing (IIH-MSP), 2012 Eighth International Conference on, 2012, pp 134-137 L Li, S Li, H Zhu, S.-C Chu, J F Roddick, and J.-S Pan, "An efficient scheme for detecting copy-move forged images by local binary patterns," Journal of Information Hiding and Multimedia Signal Processing, vol 4, pp 46-56, 2013 S.-J Ryu, M.-J Lee, and H.-K Lee, "Detection of copy-rotate-move forgery using zernike moments," in International Workshop on Information Hiding, 2010, pp 51-65 A M Moussa, "A fast and accurate algorithm for copy-move forgery detection," in Computer Engineering & Systems (ICCES), 2015 Tenth International Conference on, 2015, pp 281-285 A V Malviya and S A Ladhake, "Copy move forgery detection using low complexity feature extraction," in Electrical Computer and Electronics (UPCON), 2015 IEEE UP Section Conference on, 2015, pp 1-5 161 [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] A J Fridrich, B D Soukal, and A J Lukáš, "Detection of copy-move forgery in digital images," in in Proceedings of Digital Forensic Research Workshop, 2003 Y Cao, T Gao, L Fan, and Q Yang, "A robust detection algorithm for copymove forgery in digital images," Forensic science international, vol 214, pp 33-43, 2012 Y Sutcu, B Coskun, H T Sencar, and N Memon, "Tamper detection based on regularity of wavelet transform coefficients," in Image Processing, 2007 ICIP 2007 IEEE International Conference on, 2007, pp I-397-I-400 M K Bashar, K Noda, N Ohnishi, H Kudo, T Matsumoto, and Y Takeuchi, "Wavelet-Based Multiresolution Features for Detecting Duplications in Images," in MVA, 2007, pp 264-267 G Li, Q Wu, D Tu, and S Sun, "A sorted neighborhood approach for detecting duplicated regions in image forgeries based on DWT and SVD," in Multimedia and Expo, 2007 IEEE International Conference on, 2007, pp 17501753 E S Khan and E A Kulkarni, "An efficient method for detection of copymove forgery using discrete wavelet transform," International Journal on Computer Science and Engineering, vol 2, p 2010, 1801 O M Prathibha, Swathikumari, N S., Sushma, P., "Image forgery detection using dyadic Wavelet transform," International Journal of Electronics Signals and Systems vol 2, pp 41-43, 2012 P Yadav and Y Rathore, "Detection of copy-move forgery of images using discrete wavelet transform," International Journal on Computer Science and Engineering, vol 4, p 565, 2012 A Myna, M Venkateshmurthy, and C Patil, "Detection of region duplication forgery in digital images using wavelets and log-polar mapping," in Conference on Computational Intelligence and Multimedia Applications, 2007 International Conference on, 2007, pp 371-377 S Bayram, H T Sencar, and N Memon, "An efficient and robust method for detecting copy-move forgery," in Acoustics, Speech and Signal Processing, 2009 ICASSP 2009 IEEE International Conference on, 2009, pp 1053-1056 M Bashar, K Noda, N Ohnishi, and K Mori, "Exploring duplicated regions in natural images," IEEE Transactions on Image Processing, 2010 A Kashyap and S D Joshi, "Detection of copy-move forgery using wavelet decomposition," in Signal Processing and Communication (ICSC), 2013 International Conference on, 2013, pp 396-400 J Yang, P Ran, D Xiao, and J Tan, "Digital image forgery forensics by using undecimated dyadic wavelet transform and Zernike moments," J Comput Inf Syst, vol 9, pp 6399-6408, 2013 E Mohebbian and M Hariri, "Increase the efficiency of DCT method for detection of copy-move forgery in complex and smooth images," in Knowledge-Based Engineering and Innovation (KBEI), 2015 2nd International Conference on, 2015, pp 436-440 162 [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] R Dixit, R Naskar, and S Mishra, "Blur-invariant copy-move forgery detection technique with improved detection accuracy utilising SWT-SVD," IET Image Processing, vol 11, pp 301-309, 2017 R Dixit, R Naskar, and A Sahoo, "Copy-move forgery detection exploiting statistical image features," in Wireless Communications, Signal Processing and Networking (WiSPNET), 2017 International Conference on, 2017, pp 22772281 H Li, W Luo, X Qiu, and J Huang, "Image forgery localization via integrating tampering possibility maps," IEEE Transactions on Information Forensics and Security, vol 12, pp 1240-1252, 2017 Y Wo, K Yang, G Han, H Chen, and W Wu, "Copy–move forgery detection based on multi-radius PCET," IET Image Processing, vol 11, pp 99-108, 2016 H Huang, W Guo, and Y Zhang, "Detection of copy-move forgery in digital images using SIFT algorithm," in Computational Intelligence and Industrial Application, 2008 PACIIA'08 Pacific-Asia Workshop on, 2008, pp 272-276 I Amerini, L Ballan, R Caldelli, A Del Bimbo, and G Serra, "A sift-based forensic method for copy–move attack detection and transformation recovery," IEEE Transactions on Information Forensics and Security, vol 6, pp 10991110, 2011 J Li, X Li, B Yang, and X Sun, "Segmentation-based image copy-move forgery detection scheme," IEEE Transactions on Information Forensics and Security, vol 10, pp 507-518, 2015 E Ardizzone, A Bruno, and G Mazzola, "Copy–move forgery detection by matching triangles of keypoints," IEEE Transactions on Information Forensics and Security, vol 10, pp 2084-2094, 2015 D M Uliyan, H A Jalab, and A W A Wahab, "Copy move image forgery detection using hessian and center symmetric local binary pattern," in Open Systems (ICOS), 2015 IEEE Confernece on, 2015, pp 7-11 A Kaur and R Sharma, "Optimization of copy-move forgery detection technique," Computer Engineering and Applications Journal, vol 2, 2013 M Zandi, A Mahmoudi-Aznaveh, and A Talebpour, "Iterative copy-move forgery detection based on a new interest point detector," IEEE Transactions on Information Forensics and Security, vol 11, pp 2499-2512, 2016 C Rafael and E Richard, "Digital Image Processing.," ed: Prentice Hall, NJ, 2002 E Harzheim, "Ordered sets, volume of Advances in Mathematics," ed: Springer) Springer, New York, 2005 C Yau, "R Tutorial with Bayesian Statistics Using OpenBUGS," URL http://www r-tutor com/conten ‘dr-tutorial-ebook Indice de instrucciones, 2012 D Elena and M Michel, "Encyclopedia of Distances," ed: Springer, 2009 T K Huynh, Le,T.T., Nguyen,T.T., Huynh, K.V., "Copy-Move Forgery Detection Techniques: A survey and New Approach," in The Third Asian Conference on Information Systems (ACIS) 2014, Nha Trang, Vietnam, 2014 H Farid, "Image forgery detection," IEEE Signal processing magazine, vol 26, pp 16-25, 2009 163 [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] Y Sheng and L Shen, "Orthogonal Fourier–Mellin moments for invariant pattern recognition," JOSA A, vol 11, pp 1748-1757, 1994 C Chong, "A formulation of a new class of continuous orthogonal moment invariants, and the analysis of their computational aspects," Kuala Lumpur: Faculty of Engineering, University of Malaya, 2003 C.-W Chong, P Raveendran, and R Mukundan, "A comparative analysis of algorithms for fast computation of Zernike moments," Pattern Recognition, vol 36, pp 731-742, 2003 C.-Y Wee and R Paramesran, "On the computational aspects of Zernike moments," Image and Vision Computing, vol 25, pp 967-980, 2007 K Wagstaff, C Cardie, S Rogers, and S Schrödl, "Constrained k-means clustering with background knowledge," in ICML, 2001, pp 577-584 T Le-Tien, T Huynh-Ngoc, and T Huynh-Kha, "The total error limits in duplicated image by modifying the paremeters of Zernike Moments computation," 2015 J Wang, G Liu, Z Zhang, Y Dai, and Z Wang, "Fast and robust forensics for image region-duplication forgery," Acta Automatica Sinica, vol 35, pp 14881495, 2009 T Le-Tien, T Huynh-Ngoc, T Huynh-Kha, and L Marie, "Zernike MomentBased Approach for Detecting Duplicated Image Regions by a Modified Method to Reduce Geometrical and Numerical Errors," in International Conference on Computational Science and Its Applications, 2015, pp 458-475 S AlZubi, N Islam, and M Abbod, "Multiresolution analysis using wavelet, ridgelet, and curvelet transforms for medical image segmentation," Journal of Biomedical Imaging, vol 2011, p 4, 2011 B Toufik and N Mokhtar, The wavelet transform for image processing applications: INTECH Open Access Publisher, 2012 J Ma and G Plonka, "A review of curvelets and recent applications," IEEE Signal Processing Magazine, vol 27, pp 118-133, 2010 J.-L Starck, E J Candès, and D L Donoho, "The curvelet transform for image denoising," IEEE Transactions on image processing, vol 11, pp 670-684, 2002 T Le-Tien, T Huynh-Kha, A Tran-Hong, L Pham-Cong-Hoan, N Dey, and M Luong, "Combined Zernike Moment and Multiscale Analysis for Tamper Detection in Digital Images," Informatica (An International Journal of Computing and Informatics), indexed by Thomson Reuters, Scopus., vol 41, pp 59-70, 2017 S Y Hasan, "Study of Zernike moments using analytical Zernike polynomials," Advances in Applied Science Research, vol 3, pp 583-590, 2012 T Le-Tien, T Huynh-Kha, L Pham-Cong-Hoan, and A Tran-Hong, "Block Based Technique for Detecting Copy-Move Digital Image Forgeries: Wavelet Transform and Zernike Moments," in The Second International Conference on Electrical and Electronic Engineering, Telecommunication Engineering, and Mechatronics (EEETEM2016), 2016, p 26 164 [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] T Huynh-Kha, T Le-Tien, S Ha, and K Huynh-Van, "Improving the Computational Cost for Copied Region Detection in Forensic Images," Journal of Science and Technology: Issue on Information and Communications Technology, vol 2, pp 55-63, August 2016 T Le-Tien, T Huynh-Ngoc, and T Huynh-Kha, "The Total Error Limited by Modifying the Parameters of Zernike Moments Computation in Duplicated Images," Journal of Automation and Control Engineering, vol 3, 2015 S M Fadl and N A Semary, "A proposed accelerated image copy-move forgery detection," in Visual Communications and Image Processing Conference, 2014 IEEE, 2014, pp 253-257 R C Pandey, S K Singh, K Shukla, and R Agrawal, "Fast and robust passive copy-move forgery detection using SURF and SIFT image features," in Industrial and Information Systems (ICIIS), 2014 9th International Conference on, 2014, pp 1-6 C.-M Pun, X.-C Yuan, and X.-L Bi, "Image forgery detection using adaptive oversegmentation and feature point matching," IEEE Transactions on Information Forensics and Security, vol 10, pp 1705-1716, 2015 B Ustubioglu, G Muzaffer, G Ulutas, V Nabiyev, and M Ulutas, "A novel keypoint based forgery detection method based on local phase quantization and SIFT," in Electrical and Electronics Engineering (ELECO), 2015 9th International Conference on, 2015, pp 185-189 B Ustubıoglu, V Nabıyev, G Ulutas, and M Ulutas, "Image forgery detection using colour moments," in Telecommunications and Signal Processing (TSP), 2015 38th International Conference on, 2015, pp 540-544 S Wenchang, Z Fei, Q Bo, and L Bin, "Improving image copy-move forgery detection with particle swarm optimization techniques," China Communications, vol 13, pp 139-149, 2016 T Huynh-Kha, T Le-Tien, S Ha-Viet-Uyen, and K Huynh-Van, "The efficiency of applying DWT and feature extraction into copy-move images detection," in Advanced Technologies for Communications (ATC), 2015 International Conference on, 2015, pp 44-49 T Bridal, "Matlab toolbox," 2011 T Huynh-Kha, T Thuong-Le, S Ha, K Huynh-Van, and N Ly-Tu, "Forgery Detection and Identification of the Original and Duplicate Region in CopyMove Images," in The International Symposium on Electrical and Electronics Engineering, Ho Chi Minh, Vietnam, 2015 T Huynh-Kha, T Le-Tien, S Ha Viet Uyen, S Huynh-Thanh, and K HuynhVan, "Develop an algorithm for image forensics using feature comparison and sharpness estimation," in Recent Advances in Signal Processing, Telecommunications & Computing (SigTelCom), International Conference on, 2017, pp 82-87 T Le-Tien, T Huynh-Kha, and S Bui-Duc, "Blind image forgeries detection using detail components from the wavelet transform," in Advances in Digital Technologies: Proceedings of the 7th International Conference on Applications of Digital Information and Web Technologies 2016, 2016, p 91 165 [73] [74] [75] [76] [77] T Huynh-Kha, T Le-Tien, S Ha-Viet-Uyen, K Huynh-Van, and M Luong, "A Robust Algorithm of Forgery Detection in Copy-Move and Spliced Images," International Journal of Advanced Computer Science & Applications, indexed by Thomson Reuters and Web of Science, vol 1, pp 1-8, 2016 T Vo-Minh, T Thuong-Le, T Huynh-Kha, and L Marie, "An embedded system implemented the multiresolution-based algorithm for forensic image detection," Journal of Sai Gon University, vol (accepted), 2017 S I Kim, K C Choi, and D S Lee, "Texture classification using run difference matrix," in Ultrasonics Symposium, 1991 Proceedings., IEEE 1991, 1991, pp 1097-1100 R Cellan-Jones, "A£ 15 computer to inspire young programmers," BBC News, vol 5, 2011 B Soni, P K Das, and D M Thounaojam, "CMFD: a detailed review of block based and key feature based techniques in image copy-move forgery detection," IET Image Processing, 2017 166 ... xuất 02 giải thuật ban đầu chưa sử dụng đa phân giải, đánh giá kết định hướng cải tiến dùng đa phân giải Chương 2-? ?Phân tích đa phân giải thuật tốn ứng dụng đa phân giải giám định ảnh Copy- Move? ??:... xuất giải thuật nghiên cứu ban đầu, làm bước trung gian nhận định nhu cầu cần thiết đa phân giải, tận dụng khai thác ứng dụng đa phân giải để xây dựng giải thuật giám định cho ảnh Copy- Move, ... sánh 02 giải thuật đề xuất giải thuật liên quan 72 1.4 Kết luận Chương 73 CHƯƠNG 2: PHÂN TÍCH ĐA PHÂN GIẢI VÀ CÁC THUẬT TOÁN ỨNG DỤNG ĐA PHÂN GIẢI TRONG GIÁM ĐỊNH ẢNH COPY- MOVE

Ngày đăng: 08/08/2021, 17:37

Tài liệu cùng người dùng

  • Đang cập nhật ...

Tài liệu liên quan