This paper presents a handwritten document recognition system based on the convolutional neural network technique.In today’s world,handwritten document recognition is rapidly attaining the attention of researchers du...This paper presents a handwritten document recognition system based on the convolutional neural network technique.In today’s world,handwritten document recognition is rapidly attaining the attention of researchers due to its promising behavior as assisting technology for visually impaired users.This technology is also helpful for the automatic data entry system.In the proposed systemprepared a dataset of English language handwritten character images.The proposed system has been trained for the large set of sample data and tested on the sample images of user-defined handwritten documents.In this research,multiple experiments get very worthy recognition results.The proposed systemwill first performimage pre-processing stages to prepare data for training using a convolutional neural network.After this processing,the input document is segmented using line,word and character segmentation.The proposed system get the accuracy during the character segmentation up to 86%.Then these segmented characters are sent to a convolutional neural network for their recognition.The recognition and segmentation technique proposed in this paper is providing the most acceptable accurate results on a given dataset.The proposed work approaches to the accuracy of the result during convolutional neural network training up to 93%,and for validation that accuracy slightly decreases with 90.42%.展开更多
In this study, a numerical method was proposed to evaluate the calligraphy work called calligraphy evaluation system. Four classical chirographies of ＂Kaisho＂, ＂Gyosho＂, ＂Sousho＂ and ＂Hiragana＂, and 47 charact...In this study, a numerical method was proposed to evaluate the calligraphy work called calligraphy evaluation system. Four classical chirographies of ＂Kaisho＂, ＂Gyosho＂, ＂Sousho＂ and ＂Hiragana＂, and 47 characters for each chirography, were selected and analyzed by this system. The ＂Sumi＂ distribution of character was clarified from 12 directions and summarized into four parts of horizontal part, diagonal left part, vertical part and diagonal fight part. The character＇s contour line was converted to a signal data in order to calculate roundness index. The degree of character＇s radian was presented by roundness index. The smooth index was calculated at the same time. Additionally, width index, ＂Sumi＂ ratio, stability index also were calculated to contrast the features of each style. The main character points of four styles of ＂Kaisho＇, ＂Gyosho＂, ＂Sousho＂, ＂Hiragana＂ were extracted to compare each other, and provide a reference for learners. The learners could obtain the quantitative data to understand their work＇s characteristics. It can also be compared with other person＇s work by this system in order to improve learners＇ writing skill.展开更多
文摘This paper presents a handwritten document recognition system based on the convolutional neural network technique.In today’s world,handwritten document recognition is rapidly attaining the attention of researchers due to its promising behavior as assisting technology for visually impaired users.This technology is also helpful for the automatic data entry system.In the proposed systemprepared a dataset of English language handwritten character images.The proposed system has been trained for the large set of sample data and tested on the sample images of user-defined handwritten documents.In this research,multiple experiments get very worthy recognition results.The proposed systemwill first performimage pre-processing stages to prepare data for training using a convolutional neural network.After this processing,the input document is segmented using line,word and character segmentation.The proposed system get the accuracy during the character segmentation up to 86%.Then these segmented characters are sent to a convolutional neural network for their recognition.The recognition and segmentation technique proposed in this paper is providing the most acceptable accurate results on a given dataset.The proposed work approaches to the accuracy of the result during convolutional neural network training up to 93%,and for validation that accuracy slightly decreases with 90.42%.
文摘In this study, a numerical method was proposed to evaluate the calligraphy work called calligraphy evaluation system. Four classical chirographies of ＂Kaisho＂, ＂Gyosho＂, ＂Sousho＂ and ＂Hiragana＂, and 47 characters for each chirography, were selected and analyzed by this system. The ＂Sumi＂ distribution of character was clarified from 12 directions and summarized into four parts of horizontal part, diagonal left part, vertical part and diagonal fight part. The character＇s contour line was converted to a signal data in order to calculate roundness index. The degree of character＇s radian was presented by roundness index. The smooth index was calculated at the same time. Additionally, width index, ＂Sumi＂ ratio, stability index also were calculated to contrast the features of each style. The main character points of four styles of ＂Kaisho＇, ＂Gyosho＂, ＂Sousho＂, ＂Hiragana＂ were extracted to compare each other, and provide a reference for learners. The learners could obtain the quantitative data to understand their work＇s characteristics. It can also be compared with other person＇s work by this system in order to improve learners＇ writing skill.
