Motoi Iwata

Segmentation of Page Images Using the Area Voronoi Diagram

This paper presents a method of page segmentation based on the approximated area Voronoi diagram. The characteristics of the proposed method are as follows: (1) The Voronoi diagram enables us to obtain the candidates of boundaries of document components from page images with non-Manhattan layout and a skew. (2) The candidates are utilized to estimate the intercharacter and interline gaps without the use of domain-specific parameters to select the boundaries. From the experimental results for 128 images with non-Manhattan layout and the skew of 0°~45° as well as 98 images with Manhattan layout, we have confirmed that the method is effective for extraction of body text regions, and it is as efficient as other methods based on connected component analysis.

Digital watermarking method for tamper detection and recovery of JPEG images

Digital watermarking is a technique that embeds additional data into digital contents so that the distortion by embedding them is imperceptible, where the additional data are called watermark. One of the applications of digital watermarking is tamper detection and recovery. The digital watermarking methods for tamper detection and recovery can detect the tampered region in a tampered image and recover it. In this application, the digital watermarking methods solve the problem that digital contents can be tampered so that the falsification is not detected. In this manuscript, we propose a new method for tamper detection and recovery of JPEG images by using a RS code. The proposed method embeds the check symbols corresponding to an information-reduced original image into the LSBs of the quantized DCT coefficients in a JPEG image directly. Therefore the method completely extracts the embedded data from watermarked JPEG images. We investigated the performance of the proposed method in view of the image quality and the ability of tamper detection and recovery. The experiments of the image quality confirmed the imperceptibility of watermarked images and images used for recovery. The experiments of the ability of tamper detection and recovery confirmed the maximum size of recoverable tampered region and the difficulty of undetectable falsification.

Speech balloon and speaker association for comics and manga understanding

Comics and manga are one of the most important forms of publication and play a major role in spreading culture all over the world. In this paper we focus on balloons and their association to comic characters or more generally text and graphic links retrieval. This information is not directly encoded in the image, whether scanned or digital-born, it has to be understood according to other information present in the image. Such high level information allows new browsing experience and story understanding (e.g. dialog analysis, situation retrieval). We propose a speech balloon and comic character association method able to retrieve which character is emitting which speech balloon. The proposed method is based on geometric graph analysis and anchor point selection. This work has been evaluated over various comic book styles from the eBDtheque dataset and also a volume of the Kingdom manga series.

Reversible Watermarking Method for JPEG Images

In this letter, we propose a simple reversible watermarking method for JPEG images using the characteristics of JPEG compression. The method embeds a watermark into a JPEG image, and it extracts the watermark from the watermarked JPEG image and at the same time can recover the watermarked JPEG image to an original unwatermarked JPEG image. Moreover we investigate the number of embeddable blocks, the quality of watermarked images, and the increase of file-size by embedding a watermark.

Text-Line Extraction as Selection of Paths in the Neighbor Graph

This paper presents a new method of text-line extraction which can be applied to tilted non-rectangular pages. The method is characterized as follows. As the representation of physical structure of a page, we propose the neighbor graph which represents neighbors of connected components. The use of the area Voronoi diagram enables us to extract neighbors without predetermined parameters. Based on the neighbor graph, the task of text-line extraction is considered to be the selection of its paths appropriate as text-lines. We apply simple iterative selection of edges with local examination so as to reduce the computational cost. From experimental results for 50 pages with rectangular and non-rectangular layout, we discuss advantages and limitations of our method.

Semi-automatic Text and Graphics Extraction of Manga Using Eye Tracking Information

The popularity of storing, distributing and reading comic books electronically has made the task of comics analysis an interesting research problem. Different work have been carried out aiming at understanding their layout structure and the graphic content. However the results are still far from universally applicable, largely due to the huge variety in expression styles and page arrangement, especially in manga (Japanese comics). In this paper, we propose a comic image analysis approach using eye-tracking data recorded during manga reading sessions. As humans are extremely capable of interpreting the structured drawing content, and show different reading behaviors based on the nature of the content, their eye movements follow distinguishable patterns over text or graphic regions. Therefore, eye gaze data can add rich information to the understanding of the manga content. Experimental results show that the fixations and saccades indeed form consistent patterns among readers, and can be used for manga textual and graphical analysis.

Memory specs: an annotation system on Google Glass using document image retrieval

We present a system working on a wearable computer that can annotate and retrieve annotations for signs, posters, public displays etc. The only limitation of the system: the annotated objects need to contain at least 3-5 lines of text with fixed layout. We evaluate our system in a poster session scenario with 5 use cases: retrieving annotations, adding notes, taking pictures and recording audio and attaching them to posters. As a ubiquitous note taking and annotation application, we believe the system can bridge online and offline discussions and give the user a way to reflect on seen information by providing them a web interface to all the annotations they took together with the source material (in our case poster and paper publications), helping to organize their thoughts and improving their recall.

Share Me - A Digital Annotation Sharing Service for Paper Documents with Multiple Clients Support

In this paper we describe a novel annotation service which is capable of seamlessly linking physical and digital worlds through paper documents. Our service uses a real-time document image retrieval method called Locally Likely Arrangement Hashing (LLAH) for providing information associated with the document. By using this service digital annotations can be added to physical documents and shared with friends via mobile devices. We present the prototype implementation, and provide a discussion covering the technical details of the system.

Multichannel audio watermarking method by multiple embedding

The objective of this work is to increase the capacity of watermark information in the audio watermarking method based on amplitude modification. We increase the capacity of watermark information by embedding multiple watermarks in the different levels of audio data independently. Moreover, we aim to improve the sound quality by modifying the scaling curve of the conventional method.

DCT-OFDM Based Watermarking Scheme Robust Against Clipping, Rotation, and Scaling Attacks

The digital watermark is used for copyright protection, and the methods should have the robustness against various attacks such as JPEG compression, clipping, rotation, and scaling attacks. Based on the DCT-OFDM based watermarking method presented at IWIHC2014, we propose a method enhancing the performance in this paper. There are two ideas in the proposed method. One idea is to reduce the computational costs required for the synchronization recovery. The other idea is to encode watermark information by error correcting code and to decode it by maximum likelihood algorithm. As the result, our method has high robustness against the IHC evaluation criteria with less degradation of quality and less computational costs.