Texture-less Object Detection - PhD Thesis Proposal

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F15%3A00235522" target="_blank" >RIV/68407700:21230/15:00235522 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Texture-less Object Detection - PhD Thesis Proposal

Popis výsledku v původním jazyce

Learning, detecting and accurately localizing texture-less objects is a common requirement for applications in both personal and industrial robotics. Despite their ubiquitous presence, texture-less objects present significant challenges to contemporary methods for visual object detection and localization. In our work we aim at simultaneous detection of multiple texture-less objects with sub-linear complexity in the number of known objects, real time performance, robustness to occlusion and clutter, lowfalse detection rate, and accurate object localization. So far, we have proposed two methods. One method works with both color and depth features. It adopts the sliding window paradigm with an efficient cascade-style evaluation of each window location. The method can run in real-time, achieves the state of the art performance, and its practical relevance was demonstrated in a real robotic application. In the other proposed method, which works only with image edges, we focused on efficien

Název v anglickém jazyce

Texture-less Object Detection - PhD Thesis Proposal

Popis výsledku anglicky

Learning, detecting and accurately localizing texture-less objects is a common requirement for applications in both personal and industrial robotics. Despite their ubiquitous presence, texture-less objects present significant challenges to contemporary methods for visual object detection and localization. In our work we aim at simultaneous detection of multiple texture-less objects with sub-linear complexity in the number of known objects, real time performance, robustness to occlusion and clutter, lowfalse detection rate, and accurate object localization. So far, we have proposed two methods. One method works with both color and depth features. It adopts the sliding window paradigm with an efficient cascade-style evaluation of each window location. The method can run in real-time, achieves the state of the art performance, and its practical relevance was demonstrated in a real robotic application. In the other proposed method, which works only with image edges, we focused on efficien

Druh

V_souhrn - Souhrnná výzkumná zpráva

CEP obor

JD - Využití počítačů, robotika a její aplikace

OECD FORD obor

—

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2015

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Počet stran výsledku

40

Místo vydání

Praha

Název nakladatele resp. objednatele

Center for Machine Perception, K13133 FEE Czech Technical University

Verze

—