Color Scene Images Research Articles

Text Extraction from Scene Images through Color Image Segmentation and Statistical Distributions

This article proposes a scheme for automatic extraction of text from scene images. We proceed by applying statistical features based color image segmentation procedure to the RGB color scene image. The segmentation separates out homogenous (in terms of color and brightness) connected components (CCs) from the image. We assume these CCs include text components. So, prime intention of this article is to inspect these CCs in order to identify possible text components. Here, a number of shape based features are defined that distinguishes between text and non-text components. Further, during learning, the distribution of these features are considered independently and approximate them using parametric distribution families. Here, we apply a selection for the best fitted distribution using likelihood criterion. The class (text or non-text) distribution is the multiplication of the corresponding feature distributions. Consequently, during testing, the CC belongs to the class that produces the highest class distribution score. Our experiments are on the database of ICDAR 2011 Born Digital Dataset. We have obtained satisfactory performance in distinguishing between text and non-text.

Method for Extracting Circular Road Signs on the Basis of Scene Image Features

Road signs provide important information to guide and regulate the behavior of drivers and pedestrians to make their journeys safer and more comfortable. Therefore, we have already proposed a method for extracting and recognizing circular road signs on the basis of the color information and shape features derived from a color scene image. In the previous method, the use of lightness in images and the determination of a dynamic threshold for the binary processes have not been achieved yet. This paper proposes an algorithm to extract circular road signs on the basis of features of scene images. That is, the proposed method makes good use of lightness as well as hue and chroma. Also, the thresholds regarding the chroma and lightness in each binary process of the blue road signs are able to be automatically decided using local features in the target image. Tests on 295 images suggest that the proposed method can accurately extract circular road signs.

カラー情景画像における円形道路標識認識法

Road signs provide important information to guide and regulate the behavior of drivers and pedestrians in order to make their journeys safer and more comfortable.However,the shape features and color information displayed on circular road signs are not always the same.We propose a method for recognizing road signs on the basis of the shape features and color information derived from a color scene image.The proposed method works in three steps.First,a process already proposed in one of our previous studies detects road sign areas' color features and circularity using fuzzy reasoning.Second,the non-standard shape features of the detected road sign area are determined.Third,the values obtained from every sub-area of the image are used to enable the recognition of road signs.Our experimental results suggest that the proposed method can accurately recognize different types of circular road signs.

Information Extraction for Circular Road Signs in Color Scene Image

Information Extraction for Circular Road Signs in Color Scene Image

カラー情景画像における円形道路標識の内部領域推定

This paper describes an algorithm for estimation of internal area in the circular road signs from color scene image due to fuzzy reasoning. The algorithm works in three steps. First, regions of circular road sign in color scene image are detected. Secondly, color information is estimated due to fuzzy reasoning by using both color features and edge information. Thirdly, the resulting region is classified into three colors: red, blue and white. The experimental results indicate that the proposed algorithm is able to accurately estimate information of internal area in the circular road signs.

IJADE surveillant—an intelligent multi-resolution composite neuro-oscillatory agent-based surveillance system

Due to the rapid development of technology, especially in the field of Internet systems, there is an increasing demand both for intelligent, mobile and autonomous systems and for the usage and conveyance of multi-media information through cyberspace. In this paper, we propose an innovative, intelligent multi-agent based model, namely intelligent Java Agent Development Environment (iJADE), to provide an intelligent agent-based platform in the e-commerce environment. In addition to the facilities found in contemporary agent development platforms, which focus on the autonomy and mobility of multi-agents, iJADE provides an intelligent layer (known as the “Conscious Layer”) to support the implementation of various AI functionalities in order to produce “smart” agents. From an implementation point of view, we introduce an intelligent, multi-media processing system known as “ iJADE Surveillant”—an intelligent multi-resolution composite neuro-oscillatory agent-based surveillance system—which is based on the integration of the following modules. (a) An automatic coarse-to-fine figure-ground scene segmentation module using the Composite Neuro-Oscillatory Wavelet-based model. (b) An automatic human face detection and extraction module using an Active Contour Model with facial “landmarks” vectors. (c) Invariant human face identification based on the Elastic Graph Dynamic Link Model. To conform to the current (and future) multi-media system standards, all of iJADE Surveillant is implemented using the MPEG-7 system framework—with comprehensive Description Schemes, feature descriptors and a model framework. From an experimental point of view, a scene gallery of over 6000 color scene images is used to test the automatic scene segmentation. One hundred distinct human subjects (with over 1020 tested scenes) are used to test the intelligent human face identification. An overall correct (invariant) facial recognition rate of over 90% is attained. We hope that the implementation of the iJADE Surveillant can provide an invariant and higher-order intelligent object (pattern) encoding, searching and identification solution for future MPEG-7 applications.

カラー情景画像を用いた円形道路標識の情報抽出

カラー情景画像を用いた円形道路標識の情報抽出

SCENOGRAM — SCENE ANALYSIS USING COMPOSITENEURAL OSCILLATORY-BASED ELASTIC GRAPH MODEL

Scene analysis is so far one of the most important topics in machine vision. In this paper, we present an integrated scene analysis model, namely SCENOGRAM (Scene analysis using CompositENeural Oscillatory-based elastic GRAph Model). Basically the proposed scene analyzer is based on the integration of the composite neural oscillatory model with our elastic graph dynamic link model. The system involves: (1) multifrequency bands feature extraction scheme using Gabor filters, (2) automatic figure-ground object segmentation using a composite neural oscillatory model, and (3) object matching using an elastic graph dynamic link model. From the implementation point of view, we introduce an intelligent agent based scene analysis and object identification solution using the SCENOGRAM technology. From the experimental point of view, a scene gallery of over 6000 color scene images is used for automatic scene segmentation testing and object identification test. An overall correct invariant facial recognition rate of over 87% is attained. It is anticipated that the implementation of the SCENOGRAM can provide an invariant and higher-order intelligent object (pattern) encoding, searching and identification solution for future intelligent e-Business.

カラー情景画像からの文字列領域の抽出

カラー情景画像からの文字列領域の抽出