紋理識別觸覺傳感器的設(shè)計與實現(xiàn)
發(fā)布時間:2018-12-27 12:52
【摘要】:為了解決機器人手指在抓取物體時無法識別其表面紋理信息的問題,設(shè)計了一種基于視覺的紋理識別觸覺傳感器.該傳感器的硬件部分包括彈性體、照明裝置和攝像頭,而算法部分則是直接運用旋轉(zhuǎn)不變LBP算法將拍攝到的紋理圖像進行計算識別.為了加快傳感器的速度并拓寬其適用范圍,將算法移植到硬件電路并在FPGA上進行了驗證.實驗結(jié)果證明該傳感器不僅可以識別包括指紋、皮革、織物、紙張等在內(nèi)的12種紋理,而且可以達到24f/s幀率的實時性處理.
[Abstract]:In order to solve the problem that the robot fingers can not recognize the texture information of the surface of the object, a texture recognition tactile sensor based on vision is designed. The hardware part of the sensor includes elastic object, illumination device and camera, while the algorithm part uses rotation invariant LBP algorithm to compute and recognize the captured texture image. In order to speed up the speed of the sensor and broaden its application range, the algorithm is transplanted to the hardware circuit and verified on FPGA. The experimental results show that the sensor can not only identify 12 textures including fingerprint, leather, fabric and paper, but also achieve real-time processing of 24f/s frame rate.
【作者單位】: 清華大學(xué)微電子學(xué)研究所;
【基金】:國家自然科學(xué)基金(61327809)
【分類號】:TP212
[Abstract]:In order to solve the problem that the robot fingers can not recognize the texture information of the surface of the object, a texture recognition tactile sensor based on vision is designed. The hardware part of the sensor includes elastic object, illumination device and camera, while the algorithm part uses rotation invariant LBP algorithm to compute and recognize the captured texture image. In order to speed up the speed of the sensor and broaden its application range, the algorithm is transplanted to the hardware circuit and verified on FPGA. The experimental results show that the sensor can not only identify 12 textures including fingerprint, leather, fabric and paper, but also achieve real-time processing of 24f/s frame rate.
【作者單位】: 清華大學(xué)微電子學(xué)研究所;
【基金】:國家自然科學(xué)基金(61327809)
【分類號】:TP212
【相似文獻】
相關(guān)期刊論文 前4條
1 崔敏;;紋理圖像分析技術(shù)在表面粗糙度測量中的應(yīng)用[J];新技術(shù)新工藝;2013年08期
2 肖淑蘋;蔣加伏;;基于雙概率神經(jīng)網(wǎng)絡(luò)的紋理圖像識別[J];計算技術(shù)與自動化;2008年02期
3 姚景昭;葉玉堂;劉霖;劉娟秀;羅穎;葉涵;徐偉;李滄海;;小型液晶屏盒內(nèi)缺陷檢測中紋理的消除[J];光電工程;2012年10期
4 熊勇;路文初;劉繼忠;胡上序;;旋轉(zhuǎn)曲面變換PSO算法估計紋理圖像MRF參數(shù)[J];江南大學(xué)學(xué)報;2005年06期
相關(guān)碩士學(xué)位論文 前3條
1 林,
本文編號:2393098
本文鏈接:http://sikaile.net/kejilunwen/zidonghuakongzhilunwen/2393098.html
最近更新
教材專著