Principal-channels for One-sided Object Cutout
نویسندگان
چکیده
We introduce principal-channels for cutting out objects from an image by one-sided scribbles. We demonstrate that few scribbles, all from within the object of interest, are sufficient to mark it out. One-sided scribbles provide significantly less information than two-sided ones. Thus, it is required to maximize the use of image-information. Our approach is to first analyze the image with a large filter bank and generate a high-dimensional feature space. We then extract a set of principal-channels that discern one object from another. We show that by applying an iterative graph-cut optimization over the principal-channels, we can cut out the object of interest.
منابع مشابه
Segmentation Rectification for Video Cutout via One-Class Structured Learning
Recent works on interactive video object cutout mainly focus on designing dynamic foreground-background (FB) classifiers for segmentation propagation. However, the research on optimally removing errors from the FB classification is sparse, and the errors often accumulate rapidly, causing significant errors in the propagated frames. In this work, we take the initial steps to addressing this prob...
متن کاملA Holistic Approach for Data-Driven Object Cutout
Object cutout is a fundamental operation for image editing and manipulation, yet it is extremely challenging to automate it in real-world images, which typically contain considerable background clutter. In contrast to existing cutout methods, which are based mainly on low-level image analysis, we propose a more holistic approach, which considers the entire shape of the object of interest by lev...
متن کاملA column generation scheme for scheduling cutting machines in a clothing industry
This work addresses a scheduling problem arising from an application in a real industrial context, namely task scheduling in the cutting department of a clothing industry. In general, clothes are produced by assembling fabric cutouts. In the cutting department, fabric rolls on a line with a constant speed, and is cut by automated machines arranged along the line. Machines are all identical. A f...
متن کاملOn nest modules of matrices over division rings
Let $ m , n in mathbb{N}$, $D$ be a division ring, and $M_{m times n}(D)$ denote the bimodule of all $m times n$ matrices with entries from $D$. First, we characterize one-sided submodules of $M_{m times n}(D)$ in terms of left row reduced echelon or right column reduced echelon matrices with entries from $D$. Next, we introduce the notion of a nest module of matrices with entries from $D$. We ...
متن کاملFisher Discriminant Analysis (FDA), a supervised feature reduction method in seismic object detection
Automatic processes on seismic data using pattern recognition is one of the interesting fields in geophysical data interpretation. One part is the seismic object detection using different supervised classification methods that finally has an output as a probability cube. Object detection process starts with generating a pickset of two classes labeled as object and non-object and then selecting ...
متن کامل