Research of 6‐DOF pose estimation in stacked scenes

Teng, Qi; Yi, Jianjun; Li, Huazhong

doi:10.1002/cpe.6563

Cited by 1 publication

(1 citation statement)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Visual recognition lays the foundation for artificial intelligence science and is crucial for various applications, including space exploration, , smart vehicles, − internet of things, − advanced manufacturing, , and others. − Visual recognition technology mainly depends on the optical information collection, as well as further process and analysis through visual processing algorithms to acquire symbolic descriptions of the scenes, thereby identifying the objects or environments. However, aimed objects might exist in light-interference or dark environments, such as smoke fogs, caves, muddy water, or other surroundings that greatly disturb the collection of optical information.…”

Section: Introductionmentioning

confidence: 99%

Active Perception in Non-Visual Recognition Environments by Stretchable Tentacle Sensor Arrays

Yue

Tang

Wang

et al. 2022

ACS Appl. Mater. Interfaces

View full text Add to dashboard Cite

Smoke fog or other light-interference environments have intrinsic obstruction for visual recognition techniques to explore objects and surroundings. Alternatively, tactile perceptions, rather than visual observations, are commonly used by burrowing or deep-sea animals to communicate with environments. Bio-inspired by this natural wisdom, here, we demonstrate stretchable tentacle sensor arrays, which can recognize surrounding objects located in non-visual conditions such as smoke fog or dark environment. Each tentacle sensor is composed of two functional parts: a retractable tentacle with a magnetic top and an elastomer bottom containing copper coils. Different from traditionally passive tactile sensors, these tentacle sensors can actively stretch under the control of a syringe pump, yielding different electrical signals when in contact with the objects. Analyzing collected sensing signals of those tactile sensor arrays by the feature analysis model, complex morphological information of irregular objects in the smoke fog can be recognized. Our study reveals a fundamental connection between stretchable tactile sensors and feature analysis and demonstrates its practical potential for active perception in a non-visual recognition environment.

show abstract