2 : 3 (Myeongah Cho et al.: Three-Dimensional Rotation Angle Preprocessing and Weighted Blending for Fast Panoramic Image Method) (Special Paper) 23 2

Size: px

Start display at page:

Download "2 : 3 (Myeongah Cho et al.: Three-Dimensional Rotation Angle Preprocessing and Weighted Blending for Fast Panoramic Image Method) (Special Paper) 23 2"

솔비 임
4 years ago
Views:

1 (Special Paper) 232, (JBE Vol. 23, No. 2, March 2018) ISSN (Online) ISSN (Print) 3 a), a), a) Three-Dimensional Rotation Angle Preprocessing and Weighted Blending for Fast Panoramic Image Method Myeongah Cho a), Junsik Kim a), and Kyuheon Kim a).,. 3.. Abstract Recently panoramic image overcomes camera limited viewing angle and offers wide viewing angle by stitching plenty of images. In this paper, we propose pre-processing and post-processing algorithm which makes speed and accuracy improvements when making panoramic images. In pre-processing, we can get camera sensor information and use three-dimensional rotation angle to find RoI(Region of Interest) image. Finding RoI images can reduce time when extracting feature point. In post-processing, we propose weighted minimal error boundary cut blending algorithm to improve accuracy. This paper explains our algorithm and shows experimental results comparing with existing algorithms. Keyword : Panorama, Blending, Azimuth, Stitching a) (Department of Electronic Engineering, KyungHee University) Corresponding Author : (Kyuheon Kim) kyuheonkim@khu.ac.kr Tel: ORCID: (). (No , Ultrawide viewing ) This work was supported by Institute for Information & communications Technology Promotion(IITP) grant funded by the Korea government (MSIT) (No , Development of generation and consumption of Jigsaw-liked Ultra-Wide Viewing Spacial Media) Manuscript received January 4, 2018; Revised March 16, 2018; Accepted March 16, Copyright 2016 Korean Institute of Broadcast and Media Engineers. All rights reserved. This is an Open-Access article distributed under the terms of the Creative Commons BY-NC-ND ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited and not altered.

2 (JBE Vol. 23, No. 2, March 2018). AR, 360 VR,..,,, [1].,,. SIFT(Scale invariant feature transform) DoG(Difference of Gaussian), (discriptor) [2].,,. SURF(speed up robust feature) SIFT [3]. FAST... RoI(Region of Interest),. (Yaw), (Pitch), (Roll) RoI. (Blending).. [4]. (Laplacian) (Pyramid Blending), (Gradient) (Poissom Blending), (Minimal Error Boundary Cut Blending) ,.. 1. (Euler angle) (a) 3 (O). z xy α(yaw)

3 (a) (b) 1. 3 Fig 1. Three-dimensional rotation angle, x x β (Pitch), z γ (Roll). +. 1(b). x, y, z, [5]. 0~2π..,. π~+π.., π~+π.. (warping). (grouping) [6]. 2. (Minimal Error Boundary Cut Blending). (a) (b) 2. Fig 2. Minimal Error Boundary Cut Algorithm

(JBE Vol. 23, No. 2, March 2018) (ghosting).. (minimal cost path). 2(a), e (1).. (2) E(error weight matrix). 2(b). ( ) ( ) ( ).. [7].. 1.

4 (JBE Vol. 23, No. 2, March 2018) (ghosting).. (minimal cost path). 2(a), e (1).. (2) E(error weight matrix). 2(b). ( ) ( ) ( ).. [7] , 3.. ORB(Oriented FAST and Rotated BRIEF). (sobel) min (focal length) 3. Fig 3. Panoramic image processing flow chart

5 . 4. Fig 4. Camera cross-sectional diagram and lens viewing angle 4 F,. s,, 3,,. (3) θ.. 5(a),, A, B, 2 z. A, B x, y dx, dy.. 5(b) A, B y x. dx A B. (W) (H), (θ) (5) dx. (4) A, B, (5). tan s F y dy (a) (b)y 5. Fig 5. Two images having RoI area

(JBE Vol. 23, No. 2, March 2018). (6) A, B,. dx, dy.,. 3. ORB(Oriented FAST and Rotated BRIEF). ORB FAST (descriptor) BRIEF FAST. FAST n ( 3 ). SIFT SURF. ORB 9 FAST-9. intensity centroid (7).

6 (JBE Vol. 23, No. 2, March 2018). (6) A, B,. dx, dy.,. 3. ORB(Oriented FAST and Rotated BRIEF). ORB FAST (descriptor) BRIEF FAST. FAST n ( 3 ). SIFT SURF. ORB 9 FAST-9. intensity centroid (7). m, C (centroid) θ. (gradient) BIN, MAX. ORB FAST (Harris corner) (multi-scale). BRIEF ORB BRIEF. in-plane BRIEF [8]. ORB (Weighted Minimal Error Boundary Cut). 6. ROI ( ) Fig 6. Result of Minimal Error Boundary Cut Blending

. 6. ( ).. 7.. e E. 25. 5 10 5 50 60 100., 5 10 50 60 100. ( ). 6, 0... (8) 7.. 8. (Sobel),. 7. Fig 7.

7 . 6. ( ) e E , ( ). 6, 0... (8) (Sobel),. 7. Fig 7. Minimal error boundary cut blending algorithm 8. Fig 8. Weighted minimal error boundary cut blending block diagram

(JBE Vol. 23, No. 2, March 2018) (E)... ROI.

Preprocessing output and comparing post processing

8 (JBE Vol. 23, No. 2, March 2018) (E)... ROI. SIFT, SURF, ORB(Oriented FAST and Rotated BRIEF). SIFT (DoG), SURF Hessian. ORB FAST (orientation). s8 4.2mm, 67.8, ,,. 9(a) (a) (b) ROI ( ) 9. Fig 9. Preprocessing output and comparing post processing algorithm with original algorithm 10. Fig 10. Speed comparing with original and proposed algorithm

.. 0.047. 9(b).... 9(b).... 10. ORB, SURF, SIFT 3. SURF, 6.269 1.67 3.75.. ORB SIFT 3.6 3.

9 (b).... 9(b) ORB, SURF, SIFT 3. SURF, ORB SIFT % (ORB ) 0.289, 70%, 85% 0.392, ,, Fig 11. Panoramic output image

(JBE Vol. 23, No. 2, March 2018).,,.. 3. 3..... ROI. GPS (grouping).,., IoT. (References) [1] J. Yang, J. Jeon and G.

10 (JBE Vol. 23, No. 2, March 2018).,, ROI. GPS (grouping).,., IoT. (References) [1] J. Yang, J. Jeon and G. Park, Overlap Estimation for Panoramic Image Generation, The Journal of Korea Society of Communication and Space Technology, Vol.9, No. 4, pp , [2] S. Park, J. Kim, J. Yoo, Fast Stitching Algorithm by using Feature Tracking, Journal of Broadcast Engineering, Vol.20, No. 5, pp , [3] M. Brown, D.G. Lowe, Automatic Panoramic Image Stitching using Invarient Feature, International Journal of Computer Vision, Vol. 74, Issue.1, pp , [4] Telecommunications Technology Association, Technical Report for High Quality Panorama Video, [5] M. Kim, S. Kim, Enhancement on 3 DoF Image Stitching Using Inertia Sensor Data, Journal of Broadcasting Engineering, vol.22, no.1, pp.51-61, [6] J. Im, E. Lee, H. Kim, K. Kim, Images Grouping Technology based on Camera Sensors for Efficient Stitching of Multiple Images, Journal of Broadcasting Engineering, vol.22, no.6, pp , [7] Alexei A. Efros, William T. Freeman, Image quilting for texture synthesis and transfer, Proceedings of the 28th annual conference on Computer graphics and interactive techniques, p , August 2001 [8] Rublee E, Rabaud V, Konolige K, Bradski G, ORB: an efficient alternative to SIFT or SURF., international conference on computer vision (ICCV), : ~ : - ORCID : - :,,,

- 2017 : - 2017 ~ : - ORCID : https://orcid.

upon Tyne - 1996 7 : University of Newcastle upon Tyne - 1996 ~ 1997 :

11 : ~ : - ORCID : - :,, : : University of Newcastle upon Tyne : University of Newcastle upon Tyne ~ 1997 : University of Sheffield, Research Fellow ~ 2006 : ~ : - ORCID : - :,,

1 : 360 VR (Da-yoon Nam et al.: Color and Illumination Compensation Algorithm for 360 VR Panorama Image) (Special Paper) 24 1, (JBE Vol. 24, No

1 : 360 VR (Da-yoon Nam et al.: Color and Illumination Compensation Algorithm for 360 VR Panorama Image) (Special Paper) 24 1, (JBE Vol. 24, No 1: 360 VR (Da-yoon Nam et al.: Color and Illumination Compensation Algorithm for 360 VR Panorama Image) (Special Paper) 24 1, 2019 1 (JBE Vol. 24, No. 1, January 2019) https://doi.org/10.5909/jbe.2019.24.1.3