Character Recognition Process Research Articles

The Influence of Writing Experiences on Holistic Processing in Chinese Character Recognition

Holistic processing has been shown to be a behavioral marker of face recognition and object recognition in experts. In contrast, Hsiao and Cottrell (2009) showed that reduced holistic processing is a marker of visual expertise in Chinese character recognition. Here we tested Chinese-literates who can read and write Chinese characters (Writers) and literates whose reading performance far exceeded their writing ability (Limited-writers). We found that Writers perceived Chinese characters less holistically than Limited-writers. In addition to Hsiao and Cottrell's (2009) findings, our study further showed that such reduction in holistic processing can be explained by writing experience in Chinese. This result may be because Chinese Writers exhibit a better awareness of the orthographic components of Chinese characters than Limited-writers due to their writing experience. This study also showed that Limited-writers are better at recognizing a character embedded in a word of a familiar font than when it is alone or of an unfamiliar font, suggesting that their reading performances depend on both the context and font familiarity. This study is also the first to report on the Chinese reading population that has far poorer writing performance than reading performance.

게임 캐릭터의 아이덴티티 개선방안 : 게임 캐릭터 제작방식의 기호학적 분석 중심으로

본 논문은 캐릭터의 기호학 분석을 통해 효과적인 게임 캐릭터 디자인 방법에 대해 제안하였다. 다양한 콘텐츠의 캐릭터 중 게임 캐릭터만이 가진 특징을 알아보고, 그 특징에 Scott McCloud의 "아이콘화"의 개념을 도입하여 캐릭터와 기호학의 관계에 대해 고찰하였다. 그리고 추상화된 기호적 캐릭터와 사실적 캐릭터를 구분하여 각각의 특징과 장점, 단점을 파악하고, 디자이너의 캐릭터 제작 과정과 수용자의 캐릭터 인지과정을 단계별로 나누어 기호학적으로 해석해보았다. 그 결과 두 과정을 서로 역순으로 진행된다는 것을 알 수 있었으며 수용자의 기호학 인식과정을 고려한 캐릭터 디자인 과정이 게임 콘텐츠 전반의 특징을 더 효과적으로 전달할 수 있다는 것을 확인할 수 있었다. This paper proposed an efficient game character design method through semiotic analysis of character. It inquired into the characteristics of game character of various contents characters, added the concept of "Iconic abstraction" of Scott McCloud to its characteristics, and studied on the relation between the character and the semiotics. The semiotic character from the factual character, seized the strong and weak point of each character, and divided them in the designer's character production process and gamer's character recognition process step by step for its semiotic interpretation. As a result, two processes were reversely progressed, and the character designing process considering this gamer's semiotic recognition process is expected to be the efficient character design element that can deliver the whole characteristics of game contents.

Text binarisation for complex multi coloured textual images

Background interferences due to text-like objects make the text binarisation process difficult. This paper presents a method for text binarisation which works across the localised text region from heterogeneous colour images with multi coloured texts and complex background with text-like objects. The method also supplements the variation in font size and uneven illumination of text. In the proposed integration of edge and colour analysis (IECA) algorithm, edge and colour information are integrated to binarise the text. Background interferences due to textured background are reduced with the character size uniformity check (CUC) algorithm and the edge quadrant coverage analysis (EQCA) algorithm. Then some background patches within the segmented character box are removed and image is binarised by corner vertices colour analysis (CVCA) algorithm. The proposed method has been applied on localised region from scene and caption text images and compared with Otsu, Niblack and Kasar method and shown encouraging performance of the proposed method. It is also shown that IECA algorithm can successfully binarise the images with non-uniform illumination, complex/textured background, multi coloured texts and existence of text-like objects. The proposed IECA approach is the required preprocessing for optical character recognition (OCR) and multi media processing.

Methodology for the Evaluation of the Algorithms for Text Line Segmentation Based on Extended Binary Classification

Text line segmentation represents the key element in the optical character recognition process. Hence, testing of text line segmentation algorithms has substantial relevance. All previously proposed testing methods deal mainly with text database as a template. They are used for testing as well as for the evaluation of the text segmentation algorithm. In this manuscript, methodology for the evaluation of the algorithm for text segmentation based on extended binary classification is proposed. It is established on the various multiline text samples linked with text segmentation. Their results are distributed according to binary classification. Final result is obtained by comparative analysis of cross linked data. At the end, its suitability for different types of scripts represents its main advantage.

A Script Independent Technique for Extraction of Characters from Handwritten Word Images

A script independent character segmentation from word images technique has been reported here. Word to character segmentation is an important preprocessing step of optical character recognition process. But in case of handwritten text, presence of touching characters decreases the accuracy of the technique of the segmentation of the characters from the word. In this paper, segmentation of handwritten word of four different scripts namely, Bangla, Devanagri, Gurmukhi and Syloti are considered as the test samples. All these scripts are characterized by the presence of a distinct line along the top of the most of the characters forming the words, called the headline or Matra. Unlike English script, the characters of these handwritten scripts and its components often encircle the main character, making the conventional segmentation methodologies inapplicable. For the segmentation technique two fuzzy features, to identify the Matra region and potential segmentation point, are used here. Experimental results, using the proposed segmentation technique, on sample of 400 handwritten word images containing all the above mentioned scripts of Bangla, Devanagri, Gurmukhi and Syloti show a success rate of 95.41%, 93.61%, 91.23% and 92.37% respectively.

Learning through Hand- or Typewriting Influences Visual Recognition of New Graphic Shapes: Behavioral and Functional Imaging Evidence

Fast and accurate visual recognition of single characters is crucial for efficient reading. We explored the possible contribution of writing memory to character recognition processes. We evaluated the ability of adults to discriminate new characters from their mirror images after being taught how to produce the characters either by traditional pen-and-paper writing or with a computer keyboard. After training, we found stronger and longer lasting (several weeks) facilitation in recognizing the orientation of characters that had been written by hand compared to those typed. Functional magnetic resonance imaging recordings indicated that the response mode during learning is associated with distinct pathways during recognition of graphic shapes. Greater activity related to handwriting learning and normal letter identification was observed in several brain regions known to be involved in the execution, imagery, and observation of actions, in particular, the left Broca's area and bilateral inferior parietal lobules. Taken together, these results provide strong arguments in favor of the view that the specific movements memorized when learning how to write participate in the visual recognition of graphic shapes and letters.

A TMS examination of semantic radical combinability effects in Chinese character recognition

The proposal of human foveal splitting assumes a vertical meridian split in the foveal representation and the consequent contralateral projection of information in the two hemifields to the two hemispheres and has been shown to have important implications for visual word recognition. According to this assumption, in Chinese character recognition, the two halves of a centrally fixated character may be initially projected to and processed in different hemispheres. Here, we describe a repetitive transcranial magnetic stimulation (rTMS) investigation of hemispheric processing in Chinese character recognition, through examining semantic radical combinability effects in a character semantic judgment task. The materials used were a dominant type of Chinese character which consists of a semantic radical on the left and a phonetic radical on the right. Thus, according to the split fovea assumption, the semantic and phonetic radicals are initially projected to and processed in the right hemisphere and the left hemisphere, respectively. We show that rTMS over the left occipital cortex impaired the facilitation of semantic radicals with large combinability, whereas right occipital rTMS did not. This interaction between stimulation site and radical combinability reveals a flexible division of labor between the hemispheres in Chinese character recognition, with each hemisphere responding optimally to the information in the contralateral visual hemifield to which it has direct access. The results are also consistent with the split fovea claim, suggesting functional foveal splitting as a universal processing constraint in reading.

Caption analysis and recognition for building video indexing systems

In this paper, we propose several methods for analyzing and recognizing Chinese video captions, which constitute a very useful information source for video content. Image binarization, performed by combining a global threshold method and a window-based method, is used to obtain clearer images of characters, and a caption-tracking scheme is used to locate caption regions and detect caption changes. The separation of characters from possibly complex backgrounds is achieved by using size and color constraints and by cross examination of multiframe images. To segment individual characters, we use a dynamic split-and-merge strategy. Finally, we propose a character recognition process using a prototype classification method, supplemented by a disambiguation process using support vector machines, to improve recognition outcomes. This is followed by a postprocess that integrates multiple recognition results. The overall accuracy rate for the entire process applied to test video films is 94.11%.

Printed Japanese Character Recognition Using Multiple Commercial OCRs

This paper proposes two algorithms for maintaining matching between lines and characters in text documents output by multiple commercial optical character readers (OCRs). (1) a line matching algorithm using dynamic programming (DP) matching and (2) a character matching algorithm using character string division and standard character strings. The paper proposes a method that introduces majority logic and reject processing in character recognition. To demonstrate the feasibility of the method, we conducted experiments on line matching recognition for 127 document images using five commercial OCRs. Results demonstrated that the method extracted character areas with more accuracy than a single OCR along with appropriate line matching. The proposed method enhanced recognition from 97.61% provided by a single OCR to 98.83% in experiments using the character matching algorithm and character recognition. This method is expected to be highly useful in correcting locations at which unwanted lines or characters occur or required lines or characters disappear.

Sublexical processing in visual recognition of Chinese characters: Evidence from repetition blindness for subcharacter components

Repetition blindness (RB) refers to the failure to detect the second occurrence of a repeated item in rapid serial visual presentation (RSVP). In two experiments using RSVP, the ability to report two critical characters was found to be impaired when these two characters were identical (Experiment 1) or similar by sharing one repeated component (Experiment 2), as opposed to when they were different characters with no common components. RB for the whole character occurred when the exposure duration was more than 50 ms with one intervening character between the two critical characters (lag=1), whereas RB for subcharacter components was more evident at exposure durations shorter than 50 ms with no intervening character (lag=0). These results provide support for the model of sublexical processing in Chinese character recognition.

Effects of Ambient Light, Camcorder Settings, and Automated License Plate Reader Settings on Plate Transcription Rates

The ability of an automated license plate reading (ALPR) system to convert video images of license plates into computer records depends on many factors. Of these, two are readily controlled by the operator: the quality of the video images captured in the field and the internal settings of the ALPR used to transcribe these images. A third factor, the light conditions under which the license plate images are acquired, is less easily managed, especially when camcorders are used in the field under ambient light conditions. A set of experiments was conducted to test the effects of ambient light conditions, video camcorder adjustments, and internal ALPR settings on the percent of correct reads attained by a specific type of ALPR, one whose optical character recognition process is based on template matching. Images of rear license plates were collected under four ambient light conditions: overcast with no shadows, and full sunlight with the sun in front of the camcorder, behind the camcorder, and orthogonal to the line of sight. Three camcorder exposure settings were tested. Two of the settings made use of the camcorder’s internal light meter, and the third relied solely on operator judgment. The license plates read ranged from 41% to 72%, depending most strongly on ambient light conditions. In all cases, careful adjustment of the ALPR led to significantly improved read rates over those obtained by using the manufacturer’s recommended default settings. Exposure settings based on the operator’s judgment worked best in all instances.

An Evolutionary Coupled Neural Oscillators with Application to Pattern Recognition

Cyclical activities are basic characteristics of all living organisms. Neurobiologists have discovered that a single neuron often possesses membrane properties that are responsible for the generation of oscillations. When coupled with other neurons, oscillations with varying properties depending on the type of interconnection can be generated. Using synchronization and temporal correlation of these oscillations can cany out the tasks of pattern recognition of different objects. The speed of recognition depends on the speed of synchronization. In this paper, we propose evolutionary coupled neural oscillators to minimize the time of synchronization through the optimization of the neuron parameters by means of a genetic algorithm. The genetic algorithm, with its global search capability, finds the optimum neuron parameters through a fitness measure that reflects the correlation strength between oscillators. Thus avoiding the trial-and-error process of estimating the neuron parameters. The superiority of the method is demonstrated through an application of character recognition process.

ASA Annual Meeting Abstracts

The F. M. James III Professor of Anesthesiology; Wake Forest University School of Medicine; Medical Center Boulevard; Winston-Salem, North Carolina;eisenach@wfubmc.eduJames C. Eisenach, M.D., EditorURL: http://www.asa-abstracts.com/Institution: Anesthesiology-ASAThe ASA Annual Meeting seems to get larger and larger, and locating the particular abstracts I want to see in the large book of abstracts is a laborious process. Here is just the site to make this easier. As it states in its home page, “Under the leadership of EMIT Committee member Ira Rampil, M.D., ASA broke new ground in 1995 when it provided a searchable database of scientific abstract titles, authors and presentation times on the Internet. In 1997 this database was moved to the ASA Web site.” I thought I would give it a try before heading down to Orlando for the 1998 meeting.For the 1998 meeting, the database consists of the 1,428 accepted papers. The following disclaimer is noted in the home page, “ASA transferred the electronic papers into the Web site as submitted; ASA has not proofread them” and “A number of abstracts and/or illustrations were damaged or untranslatable. These papers are indicated in the search results.” Thus, the quality of information depends primarily on the quality of information presented in the abstracts themselves. There is also a statement that any of the electronic abstract material may not be reproduced or reprinted before presentation of the abstract at the annual meeting.Only 35% of the abstracts were submitted with a disk, and only these abstracts are available for full viewing. For the remainder of the abstracts only the titles and authors are available, and were probably scanned in, followed by a character-recognition process. Clearly, the usefulness of the search would be enhanced if all abstracts were available in complete form. I believe that electronic abstract submission for the ASA Annual Meeting will be mandatory in the immediate future, which will cure this problem.This site is under the journal Anesthesiology site, and links are available through that site, as previously reviewed. No specific links, other than to the publisher, are available through the ASA Annual Meeting Abstract site.The stated purpose of the site is to allow easy search of topics of interest for the attendee to the meeting. Abstracts can be searched by author (only the first author is included in this search), abstract number, or key word. Key words can be combined with the operators AND, OR, and NOT to refine the search. In addition, one can browse all the abstracts in a given category. The results of a search are listed with abstract title, first author, institution, abstract number, and, if available, the beginning of the full abstract. Clicking on the title reveals the full abstract, if it had been submitted electronically.Although somewhat useful, the search would be greatly enhanced if one could plan an itinerary for the actual meeting. Thus, it is not possible in the current format to browse the abstracts by session or day of the meeting, and the search results do not include date, time, and location of presentation. One needs to use the abstract number or author from the results to look up in the annual meeting book to find this information. In addition, the site does not include the entire meeting, just the scientific abstracts. A complete meeting planner on the site would also enhance its utility.Having a searchable database of abstracts for a meeting the size of the ASA Annual Meeting is a great service to the membership of the society. The capabilities of the current database are limited by low rate of electronic submissions of abstracts and the inability to plan an itinerary from the information solely available on the site. The former problem will be solved with mandatory electronic abstract submission, soon to occur for this meeting. Including the entire meeting material (abstracts, refresher course lectures, panels, and others with key words), allowing one to browse the meeting by sessions and times as well as topic, and inclusion of all times and places would make this the perfect place to optimize your experience at this meeting, which has something for everyone.James C. Eisenach, M.D.The F. M. James III Professor of Anesthesiology; Wake Forest University School of Medicine; Medical Center Boulevard; Winston-Salem, North Carolina;eisenach@wfubmc.edu(Accepted for publication September 24, 1998.)

Document retrieval tolerating character recognition errors—evaluation and application

This paper presents two methods of combining character recognition with techniques for retrieving Japanese documents and also shows how these methods can be applied to textual image retrieval. Both retrieval methods are tolerant of errors that occur during the character recognition process. The basic idea is to utilize the characteristics of recognition errors. One uses a confusion matrix to generate “equivalent” query strings that should match erroneously recognized text. The other one searches a “non-deterministic text” that contains multiple candidates for ambigous recognition results. Simulation experiments have shown that both methods can effectively combine character recognition with retrieval techniques.

Structural Indexing for Character Recognition

In this paper we present a structural method to speed up the character recognition process by reducing the number of the prototypes used during the classification of a given sample. It adopts simplified descriptions of the character shapes and uses a rough classification scheme in order to select the prototypes that most likely will match a given sample. The descriptions are stored in a multilevel data structure adopted to represent the character shape. The lowest level of such a data structure contains the detailed description of the character in terms of its skeletal features. The intermediate one consists of a list of groups of features, each one representing a character component. The upper level, eventually, is an index vector, whose dimension equals the different types of superfeatures. By using this index vector a fast and reliable selection of the prototypes to be considered as candidates for the matching can be obtained. Once this subset has been obtained, the more detailed description based on the skeletal features is resorted and the main classifier activated. Experiments have proved that the method is efficient and correct, since it allows us to select a small subset of prototypes which always contains the right one, thus reducing the classification time without affecting the accuracy of the system.

On-Line Chinese Character Recognition Via a Representation of Spatial Relationships Between Strokes

Chinese characters are constructed by basic strokes based on structural rules. In handwritten characters, the shapes of the strokes may vary to some extent, but the spatial relations and geometric configurations of the strokes are usually maintained. Therefore these spatial relations and configurations could be regarded as invariant features and could be used in the recognition of handwritten Chinese characters. In this paper, we investigate the structural knowledge in Chinese characters and propose the stroke spatial relationship representation (SSRR) to describe Chinese characters. An On-Line Chinese Character Recognition (OLCCR) method using the SSRR is also presented. With SSRR, each character is processed and is represented by an attribute graph. The process of character recognition is thereby transformed into a graph matching problem. After careful analysis, the basic spatial relationship between strokes can be characterized into five classes. A bitwise representation is adopted in the design of the data structure to reduce storage requirements and to speed up character matching. The strategy of hierarchical search in the preclassification improves the recognition speed. Basically, the attribute graph model is a generalized character representation that provides a useful and convenient representation for newly added characters in an OLCCR system with automatic learning capability. The significance of the structural approach of character recognition using spatial relationships is analyzed and is proved by experiments. Realistic testing is provided to show the effectiveness of the proposed method.

生体計測応用 注視点マスキング法による文字情報獲得機構の分析

To clarify the human mechanism that controls the relation between information processing and eye movements, we propose a fixation masking method that detects fixations and masks only those fixations. This paper describes how we examined the time required to obtain information on Japanese characters as well as the quantity of information acquired at each fixation point. Subjects were instructed to recognize incorrect characters from text they read. The results indicated that at least 67 ms were required to obtain visual information after viewing the stimulus. However, the masking affected the character recognition process in brain. From the distribution of fixation duration, approximately 130 ms were required to invoke the saccade once the subject had perceived the mask. Our results were different from Rayner's results. We assume that more complex information processing is required to read Japanese than to read English.

22-2 文字確認過程における注視点位置の解析

22-2 文字確認過程における注視点位置の解析

On-line handwritten Chinese character recognition via a fuzzy attribute representation

Chinese characters are constructed by strokes according to structural rules. Therefore, the geometric configurations of characters are important features for character recognition. In handwritten characters, stroke shapes and their spatial relations may vary to some extent. The attribute value of a structural identification is then a fuzzy quantity rather than a binary quantity. Recognizing these facts, we propose a fuzzy attribute representation (FAR) to describe the structural features of handwritten Chinese characters for an on-line Chinese character recognition (OLCCR) system. With a FAR. a fuzzy attribute graph for each handwritten character is created, and the character recognition process is thus transformed into a simple graph matching problem. This character representation and our proposed recognition method allow us to relax the constraints on stroke order and stroke connection. The graph model provides a generalized character representation that can easily incorporate newly added characters into an OLCCR system with an automatic learning capability. The fuzzy representation can describe the degree of structural deformation in handwritten characters. The character matching algorithm is designed to tolerate structural deformations to some extent. Therefore, even input characters with deformations can be recognized correctly once the reference dictionary of the recognition system has been trained using a few representative learning samples. Experimental results are provided to show the effectiveness of the proposed method.

Recognizing characters in scene images

An effective algorithm for character recognition in scene images is studied. Scene images are segmented into regions by an image segmentation method based on adaptive thresholding. Character candidate regions are detected by observing gray-level differences between adjacent regions. To ensure extraction of multisegment characters as well as single-segment characters, character pattern candidates are obtained by associating the detected regions according to their positions and gray levels. A character recognition process selects patterns with high similarities by calculating the similarities between character pattern candidates and the standard patterns in a dictionary and then comparing the similarities to the thresholds. A relaxational approach to determine character patterns updates the similarities by evaluating the interactions between categories of patterns, and finally character patterns and their recognition results are obtained. Highly promising experimental results have been obtained using the method on 100 images involving characters of different sizes and formats under uncontrolled lighting.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>