首页|Findings from Southwest Jiaotong University Reveals New Findings on Machine Lear ning (Tftsvm: Near Color Recognition of Polishing Red Lead Via Svm Based On Thre shold and Feature Transform)
Findings from Southwest Jiaotong University Reveals New Findings on Machine Lear ning (Tftsvm: Near Color Recognition of Polishing Red Lead Via Svm Based On Thre shold and Feature Transform)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News-Data detailed on Machine Learning have been presented. According to news reportingout of Chengdu, People's Republic o f China, by NewsRx editors, research stated, "With the extensiveapplication ofm achine vision in themanufacturing industry, target region recognition in complex industrialscenes is becoming a vital research territory. In the automatic poli shing of molds, polishing red lead, asan auxiliary tool for polishing positioni ng, can intuitively determine the areas to be polished."Financial support for this research came from Sichuan Science and Technology Pro gram.Our news journalists obtained a quote from the research from Southwest Jiaotong University, "Its brightcolor information are very suitable for vision-based rec ognition. Due to the interference of the near color inthe polishing environment , the traditional color recognition method has the appearance of over-segmentation. In this paper, we propose a novel near-color recognition method via SVM base d on threshold andfeature transform (TFTSVM) to improve the identification accu racy of polishing red lead. Specifically, thismethod adopts a threshold-based c olor recognition algorithm to extract two kinds of color features of redlead co lor and its near color in HSV color space and skillfully finds it is distinguish able in three dimensions.To reduce the computational complexity, a machine lear ning segmentation model is constructed, whichrealizes dimension reduction by in tegrating the feature transformation method of sample transformationand project ion transformation to achieve the best segmentation effect. Experimental results on selfestablisheddataset demonstrate that the proposed method has an excelle nt identification effect on thered lead area in the field polishing environment and also shows good robustness under the condition thatthere are reflections o n the mold surface. It meets the requirements of mechanical arm polishing and im proves the safety and reliability of automatic polishing. In addition, we also c ompare different machinelearning algorithms and advanced studies to verify the correctness of the algorithm."
ChengduPeople's Republic of ChinaAsi aCyborgsEmerging TechnologiesMachine LearningSouthwest Jiaotong University
NETL
NSTL
