This paper describes an object detection framework that learns the discriminative co-occurrence of multiple features. Feature co-occurrences are automatically found by Sequential Forward Selection at each stage of the boosting process. The selected feature co-occurrences are capable of extracting structural similarities of target objects leading to better performance. The proposed method is a generalization of the framework proposed by Viola and Jones, where each weak classifier depends only on a single feature. Experimental results obtained using four object detectors, for finding faces and three different hand gestures, respectively, show that detectors trained with the proposed algorithm yield consistently higher detection rates than those based on their framework while using the same number of features.

In this paper, we propose a new distinctive feature, called joint Haar-like feature, for detecting faces in images. This is based on co-occurrence of multiple Haar-like features. Feature co-occurrence, which captures the structural similarities within the face class, makes it possible to construct an effective classifier. The joint Haar-like feature can be calculated very fast and has robustness against addition of noise and change in illumination. A face detector is learned by stagewise selection of the joint Haar-like features using AdaBoost. A small number of distinctive features achieve both computational efficiency and accuracy. Experimental results with 5, 676 face images and 30,000 nonface images show that our detector yields higher classification performance than Viola and Jones' detector; which uses a single feature for each weak classifier. Given the same number of features, our method reduces the error by 37%. Our detector is 2.6 times as fast as Viola and Jones' detector to achieve the same performance

A new feature selection method for reliable tracking is presented. In this paper, it is assumed that features are tracked by template matching where small regions around the features are defined as templates. The proposed method selects features based on the upper bound of the average template matching error. This selection criterion is directly related to the reliability of tracking and hence, the performance is better than that of other feature detectors. Experimental results are presented to confirm the efficiency of the proposed method.

This paper proposes a new method for improving the recognition accuracy of video text by exploiting the temporal redundancy of video. The proposed method divides the video into short segments and obtains several recognition results from some video segments. The video segments have various backgrounds because background image changes temporally due to camera-work or object motion. These recognition results from diverse backgrounds are integrated into a single text string after selecting the best recognition results of individual characters. The proposed method was tested on a large set of news video sequences. Experimental results show that the proposed method increased the number of correct characters by 3.1% and the number of strings which do not include any recognition errors by 8.1%.

The paper presents a method to precisely extract only video character portions from a video text rectangle region in order to make a readable image for OCR. In conventional methods, gray image binarization processing with a given threshold is employed to extract high-intensity video character regions. A video has a complex background with various kinds of intensity so that appropriate thresholds are not always obtained. The proposed method extracts reliable high-intensity regions in a video text region and then expands them in order to make the whole video character regions. The experiments show this new method to be superior to the conventional methods.

A cut detection technique using a likelihood ratio test for MPEG compressed video sequences is proposed. Automatic cut detection techniques are essential techniques to analyze video content. The conventional techniques calculate the similarities of two successive frames and determine cut points by comparing the similarities with a given threshold value; the threshold value is determined heuristically. The new technique estimates likelihood functions from video samples and determines cut points by testing the hypothesis that two frames belong to the same shot. This technique makes it possible to calculate the optimum threshold value analytically. The experimental results show that the new technique gives the smallest probability of false negative for fixed probability of false positive in comparison with the conventional techniques.

Article Free Access Share on A shot classification method of selecting effective key-frames for video browsing Authors: Hisashi Aoki Research and Development Center, Toshiba Corporation, 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki, Kanagawa 210, Japan Research and Development Center, Toshiba Corporation, 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki, Kanagawa 210, JapanView Profile , Shigeyoshi Shimotsuji Research and Development Center, Toshiba Corporation, 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki, Kanagawa 210, Japan Research and Development Center, Toshiba Corporation, 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki, Kanagawa 210, JapanView Profile , Osamu Hori Research and Development Center, Toshiba Corporation, 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki, Kanagawa 210, Japan Research and Development Center, Toshiba Corporation, 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki, Kanagawa 210, JapanView Profile Authors Info & Claims MULTIMEDIA '96: Proceedings of the fourth ACM international conference on MultimediaFebruary 1997 Pages 1–10https://doi.org/10.1145/244130.244135Published:01 February 1997Publication History 46citation744DownloadsMetricsTotal Citations46Total Downloads744Last 12 Months34Last 6 weeks4 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF

Table form document structure analysis is an important problem in the document processing domain. The paper presents a method called Box Driven Reasoning (BDR) to robustly analyze the structure of table form documents which include touching characters and broken lines. Most previous methods employ a line oriented approach. Real documents are copied repeatedly and overlaid with printed data, resulting in characters which touch cells and lines which are broken. BDR deals with regions directly, in contrast with other previous methods. Experimental tests show that BDR reliably recognizes cells and strings in document images with touching characters and broken lines.

This paper presents a new precise line detection technique that works efficiently even for a complicated engineering drawing. Both the contours and skeletons of line regions in an image include crucial information for extracting lines precisely from a drawing. The usage of both contours and skeletons enables the following features: 1) the detected lines are hardly affected by intersecting or touching other lines; 2) the detected lines maintain the shape of the figures in the original drawing; and 3) the connectivity of lines in an image is preserved in the detected figures. Experiments on actually used drawings are discussed in order to evaluate the technique described.

A new raster-to-vector conversion method is proposed for capturing high-quality vectors from line drawing images in which the lines are overlapped by characters, symbols, or other lines. Conventionally, two raster-to-vector conversion methods have generally been used. One is a thinning method. The other is a medial line extraction method based on parallel pairs of contours. The thinning method tends to distort extructed lines, especially of intersections and corners. On the other hand, the medial line extraction method has poor capabilities as regards capturing correct lines at intersections. In contrast, the method proposed can extract precise lines from both contours and skeletons. Contours are able to hold edge shapes well, while skeletons preserve topological features; thus, a combination of these features effectively leads to the best fitting line. In the proposed method, the line which best fits the original image is selected from among various candidate lines. The candidates are created from several merged short skeleton fragments located between pairs of short contour fragments. The method is also extended to circular arc fitting. Experimental results show that the method is effective.>

This paper proposes an automatic interpretation method using probabilistic relaxation for line-drawings, which consist of lines, symbols, and characters, such as electricity distribution line diagrams superimposed on maps. An interpretation problem for lines is formulated as a labeling problem using probabilistic relaxation to obtain globally consistent results. The entire automatic interpretation consists of two stages. The first stage is segmentation and recognition for primitive components, like symbols, characters, and long lines. The second stage is long line interpretation, using probabilistic relaxation.>