近年來，隨著智慧型手機的普及，其內建的相機畫素動則高達500 萬~1300 萬，
已經逐漸取代相機，成為目前最普遍的隨身照相及錄影設備。但在使用手機拍
照或錄影時，不同於先前使用的數位相機，有時畫面會有歪斜現象產生。探討其
成因，由於目前傳統數位相機內建的感光元件大多為charge-coupled device(CCD),
採global shutter 方式擷取影像，因此不會有偏移、搖晃、部分曝光等現象；而手
機開發商由於價格的考量，多以採用捲動式快門(rolling shutter) 的complementary
metal-oxide-semiconductor(CMOS) 作為其感光元件，所以會有偏移、搖晃、部分
曝光的問題，本文僅就偏移現象加以分析並提出改善對策。
先前被提出解決偏移現象的對策，是以偏移參數及global motion vector(GMV)
對偏移的畫面自動進行復原，缺點是在某些情況下，選取GMV 時容易造成誤判。
本文提出將g-sensor 產生的加速度資料轉換為位移，g-sensor 的位移亦為相
機位移，取代GMV 所代表的位移量。改以偏移參數及相機位移直接修正rolling
shutter CMOS sensor 在相機快速移動時錄影或拍照，產生的畫面偏移現象，並重
建遺失的像素。由實驗結果得知，此演算法能有效的降低rolling shutter CMOS
sensor 造成畫面偏移現象。由於透過將加速度轉成位移即可輕易的計算出相機位
移，比起複雜GMV 的選取流程，除了可有效減輕演算法複雜度，也可排除在特
定畫面時選取GMV 可能造成的誤判

In recent years, with the smart phone popularizing, the resolution of its built-in camera
is up to 500 million~1300 million pixels, it has gradually replaced the digital camera
and become the most popular portable device which was used to take pictures or video.
However, when using a smart phone to take pictures or videos, unlike a previously used
digital camera, sometimes skew distortion occurs in the screen. The key reason is the builtin
sensors of digital camera are mostly charge-coupled device(CCD), which take global
shutter mode to capture images, so that there will be no skew/wobble/partial exposure
effect. Handset developers choose complementary metal-oxide-semiconductor(CMOS)
sensor which used rolling shutter due to cost, and it will result in the above problems. The
paper analyzed skew distortion and come up with the improvement proposal.
Solutions proposed previously based on using skew parameter and global motion
vector(GMV) to automatically recover the skew screen, the disadvantages of which are
the selected GMV is easily misjudged in some cases.
In this thesis, the accelerometer data measured by g-sensor is converted into displacement,
it is equal to camera displacement,replacing the displacement GMV represents.
Skew parameter and camera displacement was used to fix skew distortion result from
rolling shutter CMOS sensor video in a quickly-moving camera directly, and rebuild the
missing pixels. Simulation results show that the proposed method effectively reduces the
skew distortion caused by rolling shutter.
Because camera offset could be easily converted through the acceleration into the
displacement, compared to the complex process to get GMV, it can not only effectively
reduce the complexity of the algorithm but also rule out the false positives probably caused
by the selection of GMV in a particular screen.

[1] C. K. Liang, L. W. Chang, and H. H. Chen, “Analysis and compensation of rolling
shutter effect,” IEEE Transactions on Image Processing, vol. 17, pp. 1323–1330,
2008.
[2] C. Geyer, M. Meingast, and S. Sastry, “Geometric models of rolling shutter cameras,”
Omnidirectional Vision, Camera Networks and Non-classical Cameras, pp. 12–19,
October 2005.
[3] O. Ait-Aider, A. Bartoli, and N. Andreff, “Kinematics from lines in a single rolling
shutter image,” IEEE Conference on Computer Vision and Pattern Recognition,
pp. 1–6, June 2007.
[4] W. H. Cho and K. S. Hong, “Affine motion based cmos distortion analysis and cmos
digital image stabilization,” IEEE Transactions on Consumer Electronics, vol. 53(3),
pp. 833–841, August 2007.
[5] S. Baker, E. Bennett, S. B. Kang, and R. Szeliski, “Removing rolling shutter wobble,”
IEEE Conference on Computer Vision and Pattern Recognition, pp. 2392–
2399, June 2010.
[6] P. E. Forssén and E. Ringaby, “Rectifying rolling shutter video from hand-held devices,”
IEEE Conference on Computer Vision and Pattern Recognition, pp. 507–
514, 2010.
[7] D. T. Vo, S. Lertrattanapanich, and Y. T. Kim, “Automatic video deshearing for skew
sequences captured by rolling shutter cameras,” IEEE International Conference on
Image Processing, pp. 625–628, 2011.
[8] W. Hong, D. Wei, and A. U. Batur, “Video stabilization and rolling shutter distortion
reduction,” Proceedings of 2010 IEEE 17th International Conference on Image
Processing, pp. 3501–3504, 2010.
[9] D. T. Võ, J. Solé, P. Yin, C. Gomila, and T. Q. Nguyen, “Selective data pruning-based
compression using high-order edge- directed interpolation,” IEEE Transactions on
Image Processing, vol. 19, pp. 399–409, 2010.
[10] A. U. Batur and B. Flinchbaugh, “Video stabilization with optimized motion estimation
resolution,” IEEE International Conference on Image Processing, pp. 465–468,
2006.
[11] J. B. Chun, H. Jung, and C. M. Kyung, “Suppressing rollingshutter distortion of
cmos image sensors by motion vector detection,” IEEE Transactions on Consumer
Electronics, pp. 1479–1487, Nov. 2008.
[12] R. C. Gonzalez, R. E. Woods, and S. L. Eddins, Digital Processing Using MATLAB.
[13] D. Poplin, “An automatic flicker detection method for embedded camera systems,”
IEEE Transactions on Consumer Electronics, vol. 52, pp. 308–311, May 2006.