This thesis presents a robust online learning and recognition system. The basic idea is to exploit information from tracking an object during the recognition and/or learning stage to obtain increased robustness and better recognition results. Object tracking by means of an extended MSER tracker is utilized to extract local features and construct their trajectories. Compact object representation...

In this paper, a novel affine invariant descriptor for object matching is proposed. The advantage of Maximally Stable Extremal Regions (MSER) method is applied to get the most stable regions in the image. Inside each region, we pick the seeds as keypoints since MSER regions are invariant to affine transformation. Besides that, Voronoi diagram is employed to split the image into small Voronoi ce...

This thesis explores component trees, hierarchical structures from Mathematical Mor-phology, and their application to image retrieval and related tasks. The distinct com-ponent trees are analyzed and a novel classification into two superclasses is proposed,as well as a contribution to indexing and representation of the hierarchies using den-drograms. The first contribution t...

The manual Chinese word segmentation dataset WordSegCHC 1.0 which was built by eight crowdsourcing tasks conducted on the Crowdflower platform contains the manual word segmentation data of 152 Chinese sentences whose length ranges from 20 to 46 characters without punctuations. All the sentences received 200 segmentation responses in their corresponding crowdsourcing tasks and the numbers of val...

Text-detection is a well-established topic within the document recognition domain, but detecting text in natural scenes has proved to be considerably harder due to the large variability of text appearances, and due to dominating clutter. In this paper, we begin with a practical model for text appearance and use it to motivate the design of a text detection system consisting of two stages: a hyp...

An adaptive minimum symbol error rate (MSER) linear multiuser detector (MUD) is proposed for direct sequence code division multiple access (DS-CDMA) systems employing multilevel pulseamplitude modulation ( -PAM) scheme. Based on a kernel density estimation for approximating the symbol error rate (SER) from training data, a least mean squares (LMS) style stochastic gradient algorithm called the ...

In this paper, we describe a system for Android smartphones which detects and extracts Chinese text, and translates it into English. Most of the processing is done on a server; the user simply takes a picture with no other input. In this project we are mainly concerned about the text detection part, for which we use the MSER algorithm; the Optical Character Recognition (OCR) is based on Tessera...