A Handwritten Chinese Character Recognition based on Convolutional Neural Network and Median Filtering

Zhuang, Yuan; Liu, Qiong; Qiu, Chengjun; Wang, Cong; Ya, Fudong; Sabbir, Ahamed; Yan, Jiaqi

doi:10.1088/1742-6596/1820/1/012162

Cited by 12 publications

(10 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on the experimental results, the method can produce an accuracy rate of 90.91% with 5000 times training and reduce the MSE score to 0.0079. Moreover, the proposed model can perform well in real-time tests [5]. Another paper proposed the recognition of handwritten Latin characters with diacritics using CNN.…”

Section: Literature Reviewmentioning

confidence: 92%

How to Stepping up Characters Recognition using CNN Algorithm?

Khaled

2022

ijicom

View full text Add to dashboard Cite

Character recognition is very important to understand ancient culture. Various papers proposed numerous to deal with handwritten character recognition. However, several traditional remain drawbacks because methods still rely on operations based on visual capabilities. Therefore, to deal with the issue, we propose a novel recognition model using a Convolutional Neural Network to produce an effective result. To build the model, we collect datasets, do preprocessing, training with several different parameters to get the highest accuracy results. Based on experiments, our proposed model can produce an accuracy quality with a value of 98.00%.

show abstract

Section: Literature Reviewmentioning

confidence: 92%

How to Stepping up Characters Recognition using CNN Algorithm?

Khaled

2022

ijicom

View full text Add to dashboard Cite

show abstract

“…However, there are few studies on air-writing without sensors for international number identification (Table 4). For instance, the study [66] presented a sliding window-based technique for removing noise and segmenting digits from a tiny portion of the spatiotemporal input from the air-writing activity. When handling temporal data, the RNN demonstrate a substantial level of accuracy.…”

Section: Air-writing Multilingual Numeral Recognitionmentioning

confidence: 99%

A Comprehensive Literature Review on Air-written Online Handwritten Recognition

K. Jabde,

H. Patil,

D. Vibhute

et al. 2024

IJCDS

View full text Add to dashboard Cite

The COVID-19 pandemic has shifted ordinary life into a digital platform. Individuals rapidly use digital or online media to avoid the impact of touch-based platforms. In this case, touchless technology is primarily used in operation theatres, online education systems, ticket counters, etc. However, the online numerals or characters written in the air are challenging to recognize efficiently due to the writing styles of every individual. Therefore, the present research focuses on recent work on online air-written handwritten recognition. Almost eighty-plus standard articles on handwriting recognition from various journal/conference databases are considered for the current review. In addition, the generalized methodology of handwriting recognition is also provided for new research. It is observed that some online handwritten databases for some languages are freely available for research purposes. However, the online real-time air-written numeral dataset is private for Devanagari and English. Therefore, designing and developing a standard dataset for character or numeral recognition written in the air is suggested. The earlier studies achieved satisfactory results on air-writing multilingual numerals collected via or without sensors using advanced machine learning and deep learning methods. The convolutional neural network (CNN) has provided the best accuracy for several languages, excluding Devanagari.

show abstract

“…The median filter does not cause the boundary of the image to become blurred when dealing with noise, and is more suitable for dealing with random occurrences in the image, such as salt and pepper noise. 19 Three kinds of filters are used to deal with the two kinds of metal parts that need to be detected, and the denoising effect of spring-bearing seat parts and gasket parts is verified, and the most suitable filter is selected. Threshold segmentation is one of the image segmentation algorithms, also known as binarization, and its role is to extract target features from images.…”

Section: Image Preprocessing Of Double-column Metal Partsmentioning

confidence: 99%

“…Finally, the sub‐pixel edge extraction method is used to extract the inner control circle edge of the spring bearing seat part and the outer circle edge of the gasket part respectively. As shown in formula (), the edge extraction contour is fitted to a circle, and the distance between the fitted circle and the points on the contour is added and summed to minimize the distance sum 19

ε^{2} = {false\sum}_{i = 1}^{n} {((), \sqrt{{(r_{i} - α)}^{2} + {(c_{i} - β)}^{2}} goodbreak- ρ)}^{2} \to \min

…”

Section: Application Of Sub‐pixel Edge Detection In Double‐column Met...mentioning

confidence: 99%

Application of subpixel edge detection in quality control of double‐column metal parts

Xie

Chen

2023

Adv Control Appl

View full text Add to dashboard Cite

At present, most of the parts of mechanical equipment are made of metal, so the accurate detection and extraction of important characteristic parameters of metal parts is the key to determine the quality of mechanical equipment. In order to accurately extract and measure the feature size of metal parts, this research first carried out image preprocessing. Threshold segmentation is performed, and the target traits are extracted from the image to obtain the ROI. Finally, the preprocessed image is extracted from the image by the sub‐pixel edge extraction technology to extract the feature size of the circle to be measured. The difference between the measured value of the characteristic dimension of the spring bearing seat and the gasket and the real value measured by the research method is within the range of 5 and 4 μm, respectively, and both of them meet the requirements of the characteristic dimension accuracy. When the feature size of the two is repeatedly measured, the variation range of the detection results of the spring bearing seat and the gasket is within 9 and 7 μm respectively, and the detection value is much smaller than the feature size tolerance. The detection accuracy of the method can reach 95.394%. The score was 96.029 and the AUC value was 0.93, both higher than other methods. The results show that the use of sub‐pixels on metal parts The edge extraction method is extremely accurate. The research method can effectively improve the detection accuracy of the feature size of metal parts, which is of great significance to the development of the entire metal parts processing industry.

show abstract

A Handwritten Chinese Character Recognition based on Convolutional Neural Network and Median Filtering

Cited by 12 publications

References 10 publications

How to Stepping up Characters Recognition using CNN Algorithm?

How to Stepping up Characters Recognition using CNN Algorithm?

A Comprehensive Literature Review on Air-written Online Handwritten Recognition

Application of subpixel edge detection in quality control of double‐column metal parts

Contact Info

Product

Resources

About