Olarik Surinta

A Comparison of Feature and Pixel-Based Methods for Recognizing Handwritten Bangla Digits

We propose a novel handwritten character recognition method for isolated handwritten Bangla digits. A feature is introduced for such patterns, the contour angular technique. It is compared to other methods, such as the hotspot feature, the gray-level normalized character image and a basic low-resolution pixel-based method. One of the goals of this study is to explore performance differences between dedicated feature methods and the pixel-based methods. The four methods are compared with support vector machine (SVM) classifiers on the collection of handwritten Bangla digit images. The results show that the fast contour angular technique outperforms the other techniques when not very many training examples are used. The fast contour angular technique captures aspects of curvature of the handwritten image and results in much faster character classification than the gray pixel-based method. Still, this feature obtains a similar recognition compared to the gray pixel-based method when a large training set is used. In order to investigate further whether the different feature methods represent complementary aspects of shape, the effect of majority voting is explored. The results indicate that the majority voting method achieves the best recognition performance on this dataset.

Recognition of handwritten characters using local gradient feature descriptors

In this paper we propose to use local gradient feature descriptors, namely the scale invariant feature transform keypoint descriptor and the histogram of oriented gradients, for handwritten character recognition. The local gradient feature descriptors are used to extract feature vectors from the handwritten images, which are then presented to a machine learning algorithm to do the actual classification. As classifiers, the k-nearest neighbor and the support vector machine algorithms are used. We have evaluated these feature descriptors and classifiers on three different language scripts, namely Thai, Bangla, and Latin, consisting of both handwritten characters and digits. The results show that the local gradient feature descriptors significantly outperform directly using pixel intensities from the images. When the proposed feature descriptors are combined with the support vector machine, very high accuracies are obtained on the Thai handwritten datasets (character and digit), the Latin handwritten datasets (character and digit), and the Bangla handwritten digit dataset. HighlightsThis paper provides a new standard Thai handwritten character dataset for comparison of feature extraction techniques and methods.This paper propose to use local gradient feature descriptors for handwritten character recognition.This paper makes use of three complex datasets, namely Bangla, Thai, and Latin, for which very high recognition accuracies have not been obtained before.

Image Segmentation of Historical Handwriting from Palm Leaf Manuscripts

Palm leaf manuscripts were one of the earliest forms of written media and were used in Southeast Asia to store early written knowledge about subjects such as medicine, Buddhist doctrine and astrology. Therefore, historical handwritten palm leaf manuscripts are important for people who like to learn about historical documents, because we can learn more experience from them. This paper presents an image segmentation of historical handwriting from palm leaf manuscripts. The process is composed of three steps: 1) background elimination to separate text and background by Otsu’s algorithm 2) line segmentation and 3) character segmentation by histogram of image. The end result is the character’s image. The results from this research may be applied to optical character recognition (OCR) in the future.

Handwritten Character Classification using the Hotspot Feature Extraction Technique

Feature extraction techniques can be important in character recognition, because they can enhance the efficacy of recognition in comparison to featureless or pixel-based approaches. This study aims to investigate the novel feature extraction technique called the hotspot technique in order to use it for representing handwritten characters and digits. In the hotspot technique, the distance values between the closest black pixels and the hotspots in each direction are used as representation for a character. The hotspot technique is applied to three data sets including Thai handwritten characters (65 classes), Bangla numeric (10 classes), and MNIST (10 classes). The hotspot technique consists of two parameters including the number of hotspots and the number of chain code directions. The data sets are then classified by the k-Nearest Neighbors algorithm using the Euclidean distance as function for computing distances between data points. In this study, the classification rates obtained from the hotspot, mark direction, and direction of chain code techniques are compared. The results revealed that the hotspot technique provides the largest average classification rates.

Robust Face Identification with Small Sample Sizes using Bag of Words and Histogram of Oriented Gradients

Face identification under small sample conditions is currently an active research area. In a case of very few reference samples, optimally exploiting the training data to make a model which has a low generalization error is an important challenge to create a robust face identification algorithm. In this paper we propose to combine the histogram of oriented gradients (HOG) and the bag of words (BOW) approach to use few training examples for robust face identification. In this HOG-BOW method, from every image many sub-images are first randomly cropped and given to the HOG feature extractor to compute many different feature vectors. Then these feature vectors are given to a K-means clustering algorithm to compute the centroids which serve as a codebook. This codebook is used by a sliding window to compute feature vectors for all training and test images. Finally, the feature vectors are fed into an L2 support vector machine to learn a linear model that will classify the test images. To show the efficiency of our method, we also experimented with two other feature extraction algorithms: HOG and the scale invariant feature transform (SIFT). All methods are compared on two well-known face image datasets with one to three training examples per person. The experimental results show that the HOG-BOW algorithm clearly outperforms the other methods.

Comparing Local Descriptors and Bags of Visual Words to Deep Convolutional Neural Networks for Plant Recognition

The use of machine learning and computer vision methods for recognizing different plants from images has attracted lots of attention from the community. This paper aims at comparing local feature descriptors and bags of visual words with different classifiers to deep convolutional neural networks (CNNs) on three plant datasets; AgrilPlant, LeafSnap, and Folio. To achieve this, we study the use of both scratch and fine-tuned versions of the GoogleNet and the AlexNet architectures and compare them to a local feature descriptor with k-nearest neighbors and the bag of visual words with the histogram of oriented gradients combined with either support vector machines and multi-layer perceptrons. The results shows that the deep CNN methods outperform the hand-crafted features. The CNN techniques can also learn well on a relatively small dataset, Folio.

A* Path Planning for Line Segmentation of Handwritten Documents

This paper describes the use of a novel A path-planning algorithm for performing line segmentation of handwritten documents. The novelty of the proposed approach lies in the use of a smart combination of simple soft cost functions that allows an artificial agent to compute paths separating the upper and lower text fields. The use of soft cost functions enables the agent to compute near-optimal separating paths even if the upper and lower text parts are overlapping in particular places. We have performed experiments on the Saint Gall and Monk line segmentation (MLS) datasets. The experimental results show that our proposed method performs very well on the Saint Gall dataset, and also demonstrate that our algorithm is able to cope well with the much more complicated MLS dataset.

Recognizing Handwritten Characters with Local Descriptors and Bags of Visual Words

In this paper we propose the use of several feature extraction methods, which have been shown before to perform well for object recognition, for recognizing handwritten characters. These methods are the histogram of oriented gradients (HOG), a bag of visual words using pixel intensity information (BOW), and a bag of visual words using extracted HOG features (HOG-BOW). These feature extraction algorithms are compared to other well-known techniques: principal component analysis, the discrete cosine transform, and the direct use of pixel intensities. The extracted features are given to three different types of support vector machines for classification, namely a linear SVM, an SVM with the RBF kernel, and a linear SVM using L2-regularization. We have evaluated the six different feature descriptors and three SVM classifiers on three different handwritten character datasets: Bangla, Odia and MNIST. The results show that the HOG-BOW, BOW and HOG method significantly outperform the other methods. The HOG-BOW method performs best with the L2-regularized SVM and obtains very high recognition accuracies on all three datasets.

Optimization of line segmentation techniques for Thai handwritten documents

The purpose of the research is to study the optimization of line segmentation techniques for Thai handwritten documents. This research considered only single-column of Thai documents. I proposed two new techniques including comparing Thai character and sorting and distinguishing. These two techniques were used with recognized techniques on the basis of projection profile (including horizontal projection profile and stripe) in the experiment. The outcome of this research suggested that the best technique for single-column Thai documents is the new technique for sorting and distinguishing, this technique provide the accuracy of 97.11%

Comparative study between deep learning and bag of visual words for wild-animal recognition

Most research in image classification has focused on applications such as face, object, scene and character recognition. This paper examines a comparative study between deep convolutional neural networks (CNNs) and bag of visual words (BOW) variants for recognizing animals. We developed two variants of the bag of visual words (BOW and HOG-BOW) and examine the use of gray and color information as well as different spatial pooling approaches. We combined the final feature vectors extracted from these BOW variants with a regularized L2 support vector machine (L2-SVM) to distinguish between classes within our datasets. We modified existing deep CNN architectures: AlexNet and GoogleNet, by reducing the number of neurons in each layer of the fully connected layers and last inception layer for both scratch and pre-trained versions. Finally, we compared the existing CNN methods, our modified CNN architectures and the proposed BOW variants on our novel wild-animal dataset (Wild-Anim). The results show that the CNN methods significantly outperform the BOW techniques.