Planning strategy for unfolding task of clothes - isolation of clothes from washed mass

Abstract

2-2, Kand a-N is h i k ic ho Chi yo d a-k u , To k y 0, 1 0 1 J ap an AbstractThe research on the labor saving and the automation of housework by robots are proceeding rapidly with the reasons of increasing population of aged persons and advancing electronics in houses. The special problems concerning housework robots are the enormous complexity of working environment and huge number and various kinds of objects in a house. Therefore the recognition function of the robot is desired highly independent from working environment. Moreover, the robot should be able to treat various kind of "Non-solid Form Objects" exist in a house. The object of this research is to develop the technology by which the robot can manage the Non-solid form objects, especially clothes. The final goal of the research is the creation of the housekeeping robot which can manage unarranged clothes into the order at the specified place. Some of the concrete subtasks of this work are taking out one clothes, expanding, classifying, folding, and putting it on the specified place. In this paper we propose a planning strategy for unfolding task of clothes.

Visual grasp point localization, classification and state recognition in robotic manipulation of cloth: An overview

“…The related visual procedure is clearly image segmentation, and, if only one cloth item is present, this reduces just to foreground-background segmentation (or background segmentation for short). Segmentation can be as simple as a binarization to eliminate the background [28,29], or more sophisticated histogram-or graph-based segmentation procedures aiming at separating different cloth items or to get some information about its surface condition (i.e., its relief). Height -or more generally, depth-information about the cloth item is also quite valuable to determine the ideal position of the gripper for a successful grasp.…”

“…• a simplified histogram-based Ohlander's method is used for relief evaluation together with height information gained from lateral cameras to determine the exact grasp location, inserting the fingers in the ditches at both sides of a crest, in [28] (this work is expanded towards the case of textured (non-solid) cloth in [29]);…”

“…Histogram [28,29] Binarization [26] Graph-based [27] Height Lateral cam [28,29] Stereo [26,27,30,31] Range cam Region Centroid Max. dist.…”

Bottom Dressing by a Dual-Arm Robot Using a Clothing State Estimation Based on Dynamic Shape Changes

“…Unlike dressing assistance, recognition methods for clothing manipulation have been widely reported in the literature [6,19]. Because objects made of cloth are so flexible, observing their motions during manipulation is essential.…”

Robot-Aided Cloth Classification Using Depth Information and CNNs

“…The process of isolation of the cloth piece from a pile was identified as one of the first tasks to be solved for laundry manipulation in a pioneer work by Kakikura et al [1,3], where the authors separated and identified three different categories (shirt, pants, and towels) using some ad-hoc rules. The problem of grasping a unique garment from a pile has also been tackled by Monso et al [8].…”