Handwritten digit recognition: investigation of normalization and feature extraction techniques

By Chenglin Liu, Kazuki Nakashima, Hiroshi Sako, Hiromichi Fujisawa. Pattern Recognition, 2004

333 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/786015ce9b4f2436da5c9b47723b500d292ab4b4

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Xiang Bai
2. Cong Yao
3. Chenglin Liu
4. Andrew Zisserman
5. Lianwen Jin
6. Luc Van Gool
7. Yu Qiao
8. Andrea Vedaldi

Citations in the corpus

• A Method for Text Localization and Recognition in Real-World Images (2010, methodology): “The character recognition starts by normalizing the mask of the MSER to a fixed-sized matrix of 35 × 35 pixel, while retaining the centroid of the region and aspect ratio [22].”
• ARDIS: a Swedish historical handwritten digit dataset (2019, methodology): “CENPARMI dataset CENPARMI [53] is another hand-written digit dataset consisting of 6000 sample images of which 4000 (400 samples per digit class) are used for training and 2000 are used for testing.”
• High performance offline handwritten Chinese character recognition using GoogLeNet and directional feature maps (2015, methodology): “ut layer as well as the original image to construct an NNuu9 CNNinput-layer array, where NNu stands for the size of input character image. Moreover, gradient feature is proved to be effective in HCCR [20], we also apply gradient feature maps in the same way. The S obel operator is used to computer t”
• Handwritten Optical Character Recognition (OCR): A Comprehensive Systematic Literature Review (SLR) (2020, methodology): “One of such image primitive that has been used in OCR is Chain Code Histogram (CCH) [98], [99].”
• Online and offline handwritten Chinese character recognition: Benchmarking on new databases (2013, methodology): “Specifically, we use 1D and pseudo 2D normalization methods [8], gradient direction feature extraction from binary images and from grayscale images [9] [10], online stroke direction feature extraction from pen-down trajectory and from pen lifts [11] [12], classification using the modified quadratic…”