scispace - formally typeset
Search or ask a question
Author

Lochandaka Ranathunga

Bio: Lochandaka Ranathunga is an academic researcher from University of Moratuwa. The author has contributed to research in topics: Feature extraction & Histogram. The author has an hindex of 6, co-authored 68 publications receiving 152 citations. Previous affiliations of Lochandaka Ranathunga include Information Technology University.


Papers
More filters
Proceedings ArticleDOI
01 Dec 2018
TL;DR: This paper provides an approach to separate and classify objects of rice sample based on color and texture features with the help of image processing and machine learning techniques.
Abstract: Rice is one of the most consuming and important cereal grains for human being in Asian countries. In the international and national rice market, milling process is evaluated by using quality of the rice. Therefore, rice quality identification is more important. Rice quality identification is done manually by human inspectors which ensures the accuracy at some extent. But it requires a lot of man power, time consumption and judgements are subjective. Rice sample is a combination of full rice, broken rice, damaged rice, paddy, stones and foreign objects. A rice sample need to classify in to these six groups in order to identify rice quality. This paper provides an approach to separate and classify objects of rice sample based on color and texture features with the help of image processing and machine learning techniques. This method starts with image acquisition using CCD camera. Gray scale conversion, noise reduction, binarization, morphological operations are applied on the acquired images. Contours of the objects are estimated by using contour detection. Watershed algorithm is used to segmentation of touching and overlapping rice kernels. Local Binary Pattern (LBP) texture feature and color features extracted from segmented images. These features used to predict the rice sample objects using Linear Kernel based Support Vector Machine (SVM). The experiment performed on six rice categories to evaluate the suggested solution. The accuracy of segmentation and classification is 96.0% and 88.0% respectively.

14 citations

Proceedings ArticleDOI
05 Mar 2015
TL;DR: A sentiment word scoring method considering the adverb-adjective combinations of a given sentence, which is subject oriented hence this approach is proposed for a call centre domain, and it can be concluded that these enhancements have an accuracy of 74% in precision.
Abstract: Measuring sentiment strength can be considered as one of the key areas of sentiment analysis. The existing sentiment word scoring functions are based on the intensity of adjectives in the sentences. To date, there is a very minimal work has been done for measuring sentiment strength based on adverb and adjective combinations. This research proposes a sentiment word scoring method considering the adverb-adjective combinations of a given sentence. This is subject oriented hence this approach is proposed for a call centre domain. We use the linguistic analysis of adverbs of degree to satisfy the adverb-adjective scoring. Then a new approach has been introduced to calculate the sentence wise sentiment score by enhancing the existing sentence level sentiment scoring algorithm. Authors propose the way of determining the total sentiment strength of a given sentence. We describe the results of our experiments on set of 100 call centre conversation audio files between the call centre agent and customer and compare our algorithms with the existing sentiment scoring algorithms. It can be concluded that these enhancements have an accuracy of 74% in precision.

12 citations

Proceedings ArticleDOI
01 Aug 2017
TL;DR: Study presents adjacent pixel connectivity based thinning algorithm for skeletonizing handwritten characters and curvature based pattern matching and histogram formation method for recognition which can sensitively accessed the round, confusion shapes and complexity of modifier connectivity in Sinhala handwritten scripts.
Abstract: This paper presents handwritten non-cursive Sinhala character recognition method based on discrete feature extraction of the well thinned character. Study presents adjacent pixel connectivity based thinning algorithm for skeletonizing handwritten characters and curvature based pattern matching and histogram formation method for recognition which can sensitively accessed the round, confusion shapes and complexity of modifier connectivity in Sinhala handwritten scripts. The devised feature vector has shown its strength in the feature classification with greater performance. Proposed method is able to recognize 21 handwritten Sinhala characters with over 90% accuracy.

10 citations

Journal ArticleDOI
TL;DR: In this paper, the efficiency and effectiveness of the Compacted Dither Pattern Code (CDPC) combined with the Bhattacharyya classifier over MPEG-7 Dominant Colour Descriptor (DCD) was reported.
Abstract: Reduction of feature space of visual descriptors has become important due to the ‘curse of dimensionality’ problem. This paper reports the efficiency and effectiveness of the Compacted Dither Pattern Code (CDPC) combined with the Bhattacharyya classifier over MPEG-7 Dominant Colour Descriptor (DCD). Both the CDPC and DCD syntactic features use a compact feature space for colour representation. The algorithmic comparison between the two is presented in this paper, and demonstrates that there are several competitive advantages of CDPC in feature extraction and classification stages when compared to MPEG-7 DCD. The embedded texel properties, spatial colour arrangements, high compactness, and robust feature representation of CDPC have proven its effectiveness in our experimental study. Visual description experiments were conducted for ten irregular shapes-based visual concepts in videos with three setups namely CDPC with Bhattacharyya classifier, DCD without spatial coherency and DCD with spatial coherency. The visual descriptions were performed with the TRECVID 2007 development key frame dataset. The experimental results are presented in terms of three common performance measures. The results show that CDPC with Bhattacharyya classifier provides a good generalised performance for irregular shapes-based visual description as compared to the other experimental setups.

8 citations

Proceedings ArticleDOI
TL;DR: An analysis of rotational and scale invariance property of locally salient dither pattern feature with a two dimensional spatial-chromatic histogram, which expands the applicability of the visual feature.
Abstract: Compacted Dither Pattern Code (CDPC) is a recently found feature which is successful in irregular shapes based visual depiction. Locally salient dither pattern feature is an attempt to expand the capability of CDPC for both regular and irregular shape based visual depiction. This paper presents an analysis of rotational and scale invariance property of locally salient dither pattern feature with a two dimensional spatialchromatic histogram, which expands the applicability of the visual feature. Experiments were conducted to exhibit rotational and scale invariance of the feature. These experiments were conducted by combining linear Support Vector Machine (SVM) classifier to the new feature. The experimental results revealed that the locally salient dither pattern feature with the spatialchromatic histogram is rotationally and scale invariant.

8 citations


Cited by
More filters
01 Jan 2001
TL;DR: This paper builds a Java retrieval framework to compare shape retrieval using FDs derived from different signatures, and examines common issues and techniques for shape representation and normalization.
Abstract: Shape is one of the most important features in Content Based Image Retrieval (CBIR). Many shape representations and retrieval methods exists. However, most of those methods either do not well represent shape or are difficult to do normalization (making matching hard). Among them, methods based Fourier descriptors (FD) achieve both well representation and well normalization. Different shape signatures have been exploited to derive FDs, however, FDs derived from different signatures can have significant different effect on the result of retrieval. In this paper, we build a Java retrieval framework to compare shape retrieval using FDs derived from different signatures. Common issues and techniques for shape representation and normalization are also analyzed in the paper. Data is given to show the retrieval result.

221 citations

01 Jan 2009
TL;DR: In this paper, a real-time traffic light recognition system for on-vehicle camera applications is presented, which is mainly based on a spot detection algorithm and is able to detect lights from a high distance with the main advantage of being not so sensitive to motion blur and illumination variations.
Abstract: This paper introduces a new real-time traffic light recognition system for on-vehicle camera applications. This approach has been tested with good results in urban scenes. Thanks to the use of our generic “Adaptive Templates” it would be possible to recognize different kinds of traffic lights from various countries. This approach is mainly based on a spot detection algorithm therefore able to detect lights from a high distance with the main advantage of being not so sensitive to motion blur and illumination variations. The detected spots together with other shape analysis form strong hypothesis we feed our Adaptive Templates Matcher with. Even though it is still in progress, our system was validated in real conditions in our prototype vehicle and also using registered video sequences. The authors noticed a high rate of correctly recognized traffic lights and very few false alarms. Processing is performed in real-time on 640x480 images using a 2.9GHz single core desktop computer.

154 citations

Reference EntryDOI
15 Aug 2006

149 citations

Journal ArticleDOI
TL;DR: This paper proposes a novel approach that goes deeper in the classification of texts collected from Twitter and classifies these texts into multiple sentiment classes, and proves to be very accurate in binary classification and ternary classification.
Abstract: Sentiment analysis and opinion mining in social networks present nowadays a hot topic of research. However, most of the state of the art works and researches on the automatic sentiment analysis and opinion mining of texts collected from social networks and microblogging websites are oriented toward the binary classification (i.e., classification into “positive” and “negative”) or the ternary classification (i.e., classification into “positive,” “negative,” and “neutral”) of texts. In this paper, we propose a novel approach that, in addition to the aforementioned tasks of binary and ternary classifications, goes deeper in the classification of texts collected from Twitter and classifies these texts into multiple sentiment classes. While in this paper, we limit our scope to seven different sentiment classes, the proposed approach is scalable and can be run to classify texts into more classes. We first introduce SENTA, our tool built to help users select out of a wide variety of features the ones that fit the most for their application, to run the classification, through an easy-to-use graphical user interface. We then use SENTA to run our own experiments of multi-class classification. Our experiments show that the proposed approach can reach up to 60.2% accuracy on the multi-class classification. Nevertheless, the approach proves to be very accurate in binary classification and ternary classification: in the former case, we reach an accuracy of 81.3% for the same data set used after removing neutral tweets, and in the latter case, we reached an accuracy of classification equal to 70.1%.

127 citations