Furniture board segmentation based on multiple sensors and integer programming model

Abstract

In the current furniture production line, the level of automation in the stage of loading and unloading is not high enough. In order to improve its automation, a novel integer programming based method for automatically segmenting board is proposed and a multi-sensor configuration is given. In such a configuration, we include multiple cameras and Lidar sensors. The cameras attached on each board are used to collect quick response (QR) code information, while Lidar sensors can obtain each board’s contours information. We then formulate each board’s segmentation as the integer programming problem. The experimental results show that our method can achieve a very high segmentation accuracy of 95% on average.

References

[1] L PingL I, ChenQ Y, DuanX Y, ChenH, DengZ W, LiaoF. Mechanical & Electrical Engineering Technology, 2016, 45: 4(in Chinese)

[2] Gao-DeWU, Jun-HuaC, Jian-MnZ, Jian-QiangMA, Nian-FuL, FuZ. Transducer & Microsystem Technologies, 2017, 36: 157(in Chinese)

[3] LIB, LuoB. Transducer & Microsystem Technologies, 2016, 35: 150(in Chinese)

[4] LimaJ, SimõesF, UchiyamaH, TeichriebV, MarchandE. Machine Vision & Applications, 2016, 27: 1

[5] AchantaR, ShajiA, SmithK, LucchiA, FuaP, SüsstrunkS. SLIC superpixels, 2010, Switzerland, EPFL

[6] Roy A. and Todorovic S., A Multi-scale CNN for Affordance Segmentation in RGB Images, European Conference on Computer Vision, Springer International Publishing, 186 (2016).

[7] Wang J., Wang Z., Tao D., See S. and Wang G., Learning Common and Specific Features for RGB-D Semantic Segmentation with Deconvolutional Networks, European Conference on Computer Vision, Springer International Publishing, 664 (2016).

[8] PabloA, MichaelM, CharlessF, JitendraM. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33: 898

[9] FangX, HongfuW. Computer Simulation, 2011, 28: 280(in Chinese)

[10] WANGJ, ZHENGY-G, PANZ-K, ZHANGW-Z, WANGG-D. Optoelectronics Letters, 2015, 11: 307

[11] GENGH-Q, ZHANGH, XUEY-B, ZHOUM, XUG-P, GAOZ. Optoelectronics Letters, 2017, 13: 381

[12] WANGX-P, ZHANGW, CUIY. Optoelectronics Letters, 2015, 11: 395

[13] LiuW, AnguelovD, ErhanD, SzegedyC, ReedS, FuCY. SSD: Single Shot MultiBox Detector, 2015, 21

[14] https://doi.org/libdmtx.wikidot.com/general-instructions.

[15] AlvesC, ClautiauxF, CarvalhoJVD, RietzJ. Linear and Integer Programming, Dual-Feasible Functions for Integer Programming and Combinatorial Optimization, 2016, 1065

[16] Basu A., Conforti M., Summa M.D. and Paat J., The Structure of the Infinite Models in Integer Programming, International Conference on Integer Programming and Combinatorial Optimization, Springer, Cham, 63 (2017).

[17] XiaolingS, DuanL. Integer Programming, 2011, (in Chinese)

[18] JiaY, ShelhamerE, DonahueJ, KarayevS, LongJ, GirshickR. Caffe: Convolutional Architecture for Fast Feature Embedding, 2014, 675