e-journal

Learning to Segment and Track in RGBD

---

We consider the problem of segmenting and tracking deformable objects in color video with depth (RGBD) data available from commodity sensors such as the Asus Xtion Pro Live or Microsoft Kinect. We frame this problem with very few assumptions—no prior object model, no stationary sensor, and no prior 3-D map—thus making a solution potentially useful for a large number of applications, including semi-supervised learning, 3-D model capture, and object recognition. Our approach makes use of a rich feature set, including local image appearance, depth discontinuities, optical flow, and surface normals to inform the segmentation decision in a conditional random field model. In contrast to previous work in this field, the proposed method learns how to best make use of these features from ground-truth
segmented sequences. We provide qualitative and quantitative analyses which demonstrate substantial improvement over the state of the art. This paper is an extended version of our previous work. Building on our previous work, we show that it is possible to achieve an order of magnitude speedup and thus real-time performance ( 20 FPS) on a laptop computer by applying simple algorithmic optimizations to the original work. This speedup comes at only a minor cost in overall accuracy and thus makes this
approach applicable to a broader range of tasks. We demonstrate one such task: real-time, online, interactive segmentation to efficiently collect training data for an off-the-shelf object detector.
Note to Practitioners—The original motivation for this work derives from object recognition in autonomous driving, where it is desirable to identify objects such as cars and bicyclists in natural
street scenes.We showed in previous work, so long as one can segment and track objects in advance of knowing what they are, it is possible to train accurate object detectors using a small number
of hand-labeled examples combined with a large number of unlabeled examples. The key dependency of a model-free segmentation and tracking method was available because of the structure in the autonomous driving problem. This paper aims to make these techniques applicable in more general environments. In the near term, themethods presented here can be used for real-time, online, interactive object segmentation. This can ease the process of collecting training data for existing object recognition systems used in automation today. In the long term, improved implementations could be an integral part of semi-supervised object recognition systems which require few hand-labeled training examples and can produce accurate recognition results.
Index Terms—Image segmentation, machine vision, object recognition, object segmentation.

---

Ketersediaan

Tidak ada salinan data

Informasi Detail

Judul Seri

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING

No. Panggil

-

Penerbit

New York : IEEE., 2013

Deskripsi Fisik

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING, VOL. 10, NO. 4, OCTOBER 2013

Bahasa

English

ISBN/ISSN

1545-5955

Klasifikasi

-

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

VOL. 10, NO. 4, OCTOBER 2013

Subjek

TEKNIK

Info Detail Spesifik

-

Pernyataan Tanggungjawab

Yuli/Agus

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar