Re: [請益] 影像處理去背景

看板Master_D作者CCY0927 (只是個暱稱罷了)時間16年前 (2010/02/12 04:04)推噓4(4推 0噓 0→)

留言4則, 4人參與討論串2/2 (看更多)

你用 image differencing 方法當然得不到好的結果，因為這個方法雖然直覺、簡單，但最大的限制就是你的前景(移動物)必需要一定移動，才能偵測得到；倘若前景物體停止動作，馬上就會被歸到背景當中。在這個方向你要找的關鍵字是 background subtraction/background modelling/ foreground detection/moving object detection，底下兩篇是這領域少見的 survey paper，你可以參考一下。 Piccardi, M.: Background subtraction techniques: a review. In: Systems, Man and Cybernetics, 2004 IEEE International Conference on. Volume 4. ( 2004) 3099–3104 vol.4 Hall, D., Nascimento, J., Ribeiro, P., Andrade, E., Moreno, P., Pesnel, S. , List, T., Emonet, R., Fisher, R., Santos Victor, J., Crowley, J.L.: Comparison of target detection algorithms using adaptive background models . In: International workshop on Performance evaluation of Tracking and Surveillance, Beijing, China (2005) 不過既然你都說這不是你發展的重點了，那你其實 OpenCV 裡的範例抄一抄就好了。 OpenCV 實作了底下這幾篇論文： KaewTraKulPong, P., Bowden, R.: An improved adaptive background mixture model for real-time tracking with shadow detection. In: 2nd European Workshop on Advanced Video-based Surveillance Systems. (2001) Li, L., Huang, W., Gu, I. Y., and Tian, Q. 2003. Foreground object detection from videos containing complex background. In Proceedings of the Eleventh ACM international Conference on Multimedia (Berkeley, CA, USA, November 02 - 08, 2003). MULTIMEDIA '03. ACM, New York, NY, 2-10. DOI= http://doi.acm.org/10.1145/957013.957017 K. Kim, T. H. Chalidabhongse, D. Harwood and L. Davis, "Real-time Foreground-Background Segmentation using Codebook Model", Real-time Imaging, Volume 11, Issue 3, Pages 167-256, June 2005. 你如果對這領域有興趣，想多了解一點，可以接著往下看；沒有的話，在此停住就可以了。 ----------------------------------------------------------------------------- 在過去的研究，大致上來說可以分成下面幾個方向：一、Pixel-based 方式這一類的方法都是從影像中 pixel 的角度出發，只是採用不同的特性來做 model，接著利用統計方式去建立模型。這類方法的好處是直覺，壞處是容易受到雜訊影響。 Stauer, C., Grimson, W.: Adaptive background mixture models for real-time tracking. In: Computer Vision and Pattern Recognition, 1999. IEEE Computer Society Conference on. Volume 2. (1999) –252 Vol. 2 Zivkovic, Z.: Improved adaptive gaussian mixture model for background subtraction. In: ICPR (2). (2004) 28–318. Lee, D.S.: Eective gaussian mixture learning for video background subtraction. Pattern Analysis and Machine Intelligence, IEEE Transactions on 27(5) (2005) 827–832 Li, L., Huang, W., Gu, I.Y.H., Tian, Q.: Statistical modeling of complex backgrounds for foreground object detection. Image Processing, IEEE Transactions on 13(11) (2004) 1459–1472 Lee, D.S., Hull, J., Erol, B.: A bayesian framework for gaussian mixture background modeling. In: Image Processing, 2003. ICIP 2003. Proceedings. 2003 International Conference on. Volume 3. (2003) III–973–6 vol.2 Sheikh, Y., Shah, M.: Bayesian modeling of dynamic scenes for object detection. Pattern Analysis and Machine Intelligence, IEEE Transactions on 27(11) (2005) 1778–1792 Tuzel, O., Porikli, F., Meer, P.: A bayesian approach to background modeling. In: CVPR ’05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05) - Workshops, Washington, DC, USA, IEEE Computer Society (2005) 58 二、Block-based 方式這一類的方法，不再使用 pixnel-based 的角度，改用 region-based 的方式去考量。好處是因為一個一個 block 去考慮，比較不容易受到雜訊影響；但壞處是，你若沒進行後處理，得到的會是相當粗略的結果。 Huang, S.S., Fu, L.C., Hsiao, P.Y.: A region-based background modeling and subtraction using partial directed hausdor distance. In: Robotics and Automation, 2004. Proceedings. ICRA ’04. 2004 IEEE International Conference on. Volume 1. (2004) 956–960 Vol.1 Russell, D., Gong, S.: A highly ecient block-based dynamic background model. In: Proceedings. IEEE Conference on Advanced Video and Signal Based Surveillance, 2005. (2005) 417–422 Li, L., Leung, M.: Integrating intensity and texture dierences for robust change detection. Image Processing, IEEE Transactions on 11(2) ( 2002) 105–112 Zhu, Q., Avidan, S., Cheng, K.T.: Learning a sparse, corner-based representation for time-varying background modeling. In: ICCV ’05: Proceedings of the Tenth IEEE International Conference on Computer Vision (ICCV’05) Volume 1, Washington, DC, USA, IEEE Computer Society (2005) 678 –685 Heikkila, M.: A texture-based method for modeling the background and detecting moving objects. IEEE Trans. Pattern Anal. Mach. Intell. 28(4) ( 2006) 657 Senior Member-Matti Pietikainen. Liu, Y., Yao, H., Gao, W., Chen, X., Zhao, D.: Nonparametric background generation. In: Pattern Recognition, 2006. ICPR 2006. 18th International Conference on. Volume 4. (2006) 916–919 三、Hierarchical Integration 這個算是整合型研究。有研究學者開始思考，若上述方法各有其優點，那我是不是可以整合兩者來達到更好的效果呢？ Toyama, K., Krumm, J., Brumitt, B., Meyers, B.: Wallower: principles and practice of background maintenance. In: Computer Vision, 1999. The Proceedings of the Seventh IEEE International Conference on. Volume 1. (1999) 255–261 vol.1 Javed, O., Shaque, K., Shah, M.: A hierarchical approach to robust background subtraction using color and gradient information. In: Motion and Video Computing, 2002. Proceedings. Workshop on. (2002) 22–27 Cristani, M., Bicego, M., Murino, V.: Integrated region- and pixel-based approach to background modelling. In: Motion and Video Computing, 2002. Proceedings. Workshop on. (2002) 3–8 Yu-Ting Chen, Chu-Song Chen, Chun-Rong Huang and Yi-Ping Hung, "Efficient Hierarchical Method for Background Subtraction," Pattern Recognition, volume 40, number 10, pages 2706-2715, October 2007. ※ 引述《clanguage (C語言)》之銘言： : 請問一下影像處理當中, : 有沒有哪些方法具有高品質的去背景方式呢？ : 希望可以推薦一些值得參考的文獻 : 我用相減法一直難以得到好的結果 : 因為這部分不是主要要研究的那部分, : 希望能有前輩提點一下, : 不然怕一頭栽進去又卡了很久 QQ" -- ※ 發信站: 批踢踢實業坊(ptt.cc) ◆ From: 122.116.111.182 ※ 編輯: CCY0927 來自: 122.116.111.182 (02/12 04:47)

推

ujj

02/12 11:06, , 1^F

02/12 11:06, 1^F

推

DennisKao

02/12 12:22, , 2^F

02/12 12:22, 2^F

推

clanguage

02/12 16:12, , 3^F

02/12 16:12, 3^F