A Novel Approach to Detect Pedestrian from Still Images Using Random Subspace Method. (2016)

• Record Type: Journal Article
• Authors: Priya, V.V.; Rekha, P.; Reshmi, K.C.; kumar, S. Manoj; Indhulekha, K.
• Is Part Of: Procedia technology. Volume 25(2016)
• Page Start: 333
• Page End: 340
• Publication Date: 2016
• Subjects: Occlusion -- Ensemble Classifier -- Support Vector Machines -- GLCM -- DCT -- DWT -- Random Subspace Method ;; Technology -- Congresses; Technology -- Periodicals; Engineering -- Congresses; Engineering -- Periodicals; Engineering; Technology; Conference proceedings; Periodicals; 605
• DOI: 10.1016/j.protcy.2016.08.115
• Languages: English
• ISSNs: 2212-0173

Abstract

Abstract: Pedestrian detection from still images is a terribly troublesome task. Human detection is the crucial part within the systems of humanistic image reclamation, visual scrutiny, pedestrian detection, and posture recognition, home automation, robot sensing. Detecting humans is a stimulating task due to major difficulties scrolling back from the wide variability of the target, like the form, wear or pose; and thereafter the external factors, like situation, illumination, and partial occlusions. This paper detects the humans using Random Subspace Method (RSM). The detection process is only in the still images no motion information is used. By using random subspace method detects the pedestrians. To implement these using mainly three types of datasets PobleSec, INRIA and Daimler Multicue dataset, additionally used linear SVM for classification.