Cellular Neural Network for Associative Memory and Its Application to Braille Image Recognition

Abstract

Braille is widely used as communication tools for sight-impaired people. A recognition system of Braille characters is essential for those who can't read them. On the other hand, it is well-known that Cellular Neural Network for associative memory(CNN) is effective for pattern recognition, and various applications have been reported. This paper proposes an improved designing method of neighborhood, and use the Braille recognition system using CNN. We demonstrated an usefulness of the proposed system in recognition experiments. As a result, we could obtain a good recognition rate(87.9%).

Solar Irradiance Prediction with Machine Learning Algorithms: A Brazilian Case Study on Photovoltaic Electricity Generation

“…ANN algorithms are highly inspired by the sophisticated functioning of human brains, in which information is processed in parallel by billions of interconnected neurons. Neural networks have been applied to several contexts, such as problems with image identification [31], predictability of financial services [32], and even for pattern recognition in DNA [33]. This versatility enables neural networks to solve classification, clustering, or prediction (regression) problems [34].…”

A Review on Software Algorithms for Optical Recognition of Embossed Braille Characters

“…On the basis of this criterion the protrusion and depression areas are identified. An effort to recognize Braille image recognition using Cellular Neural Networks (CNN) was presented by Namba et al in 2006 [20]. In this work CNN is used as constructing element of the system where the CNN for associative memory was used for Braille cell recognition.…”

Optical Braille recognition with Haar wavelet features and Support-Vector Machine

“…We determine the coordinate of the first Braille character (x, ,y,) in the image as the minimum of coordinate of all the dots found in the image: (4) Ys = min(Yi I J(i) = 1) (5) where (Xi' yJ is the coordinate of the i-th pixel in the image, and J( i) is the intensity of the pixel.…”

“…More recently, reference [3] proposed a phone-camera-978-1-4244-7956-6/1 0/$26.00 ©20 10 IEEE 64 Dayong Zhang 877 Shen Jia Nong Road, Shanghai, China brian.zhang78@gmail.com based Braille recognition system in which a Java program can process the image taken by phone camera, and translate the Braille dots into Japanese. Reference [5] proposed the use of cellular neural network for Braille recognition, and obtained recognition rate of 87.9%.…”