Iping Supriana

Traffic sign recognition with Color-based Method, shape-arc estimation and SVM

Traffic sign recognition is studied as one of the modern assistance in driving. The main purpose of it is to help driver on realizing traffic sign around the car by using computer vision enhanced to the car. In this paper, a combination method of Color-based Method and SVM is presented to do the traffic sign recognition. Color-based Method with CIELab + hue is chosen because it gives good result on localizing traffic signs. It is first used to preprocess the image to binary image. The binary image is then processed with canny to give more accurate result on detecting traffic signs. To get the traffic sign shape, the preprocessed image is checked by using Ramer-Douglas-Peucker algorithm, this algorithm will approximate each closed object shape in the preprocessed image. Detecting closed object make it unable to detect occluded and attached road signs. In order to detect occluded and attached signs, the proposed method will use two phase of detection by estimating the shape of the arc or later called the shape-arc algorithm. Approximated circle, square, or triangle images will be marked as traffic sign and processed in the recognition step with linear c-SVM. SVM classification will be based on binary images which gives 97% accuracy. The proposed method is more accurate than methods proposed by the reference papers and the detection is able to detect harder problems such as attached signs.

Collaboration model of software development

Software development is a series of processes activity involving more than one person, especially in medium to large software projects. Each activity of software development is carried out by various combination of different individual. Collaboration occurs in every phase of the process software development. The right team collaboration method is one of the important things in success software development. This research is focused to develop collaboration model of software development to meet time schedule, budget, scope, and software quality.

Histogram based color pattern identification of multiclass fruit using feature selection

Color histogram has been widely used in feature extraction to represent color feature of an object in the image. In this paper, we identify which features that give high contribution in classification performance, because not all features are directly correlated with object category. In the case of n-bins color histogram, features were referred to color intensity range of color histogram. On the one hand, we consider fruit classification, where the feature space contains various properties of pixel intensities of RGB (Red-Green-Blue) channel. On selecting feature subset, we consider filter method of feature selection. In the filter method, we successively reduce the size of the feature sets and investigate the changes in the classification results. Specifically, we followed the filtering approach to feature selection: selecting features in a single pass first and then applying a classification algorithm independently. We used chi square feature selection to determine relevant features from RGB histogram. Further, we used and evaluated those relevant features in a classification system, using K-Nearest Neighbor (KNN) as classifier. In this paper we show that by conducting feature selection techniques combined with KNN we would be able to prune non-relevant intensities value of Red, Green, and Blue channel. Furthermore, we use the relevant subset of features to identify intensities range of RGB channel that was needed to represent 32 subcategories fruit image efficiently.

Hand detection for virtual touchpad

There are many different kinds of pointing devices for interaction between human and computer. One of the most used and developed pointing devices is the mouse. However, traditional mouse has a limited degree of controls. Another device is the touchpad, the counterpart of the mouse that is usually embedded on laptops. Touchpad basically has the same control as the mouse, but it can have additional controls because it has a different interaction method compared to the mouse. In this paper we discuss about virtual touchpad, a touchpad built using a webcam and some existing computer vision techniques, mostly hand detection. It uses a piece of rectangular object as the pad and works just like a touchpad with a few additional controls. The virtual touchpad built supports clicking (i.e. left click, right click and middle click) and gestures recognition by using manipulation of hand and finger position. The gestures recognized by this touchpad are simple gestures e.g. making a vertical line, horizontal line, V shape, and so on. Although the virtual touchpad is still a prototype, most of the user who tested this device finds it interesting and said that it would be useful if developed further.

Knowledge sharing between similar domain knowledge management systems

Knowledge Management System (KMS) provides a library of knowledge, starting from collecting, organizing, summarizing, analyzing, synthesizing, and decision-making. Each KMS usually designed to accommodate specific knowledge domain based on the expertise of the institution or company, which make each KMS has its own expertise. This characteristic of KMS gives benefit to user where they can find the perfect (deep and thorough) knowledge they need, however, this also create gap where user must first find the correct KMS and then search the correct document. This research proposes a model where a KMS has the ability to share its content (knowledge sharing) to other KMS with similar knowledge domain, which in return provides benefit such as knowledge enhancement for the first KMS (harvester). The model uses OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting) as a basis of development to find a suitable KMS (repository), retrieve specific knowledge from the KMS and proposing a mechanism to store...

Structural offline handwriting character recognition using levenshtein distance

Research in offline handwriting recognition for unconstrained text remains a difficult challenge. Some problems such as noise in image, skew of text, cursive letters, and various handwriting style is still an open problem. Many method has been researched to solve those problems, such as k-NN, Neural Network, SVM and HMM. And to improve the recognition result, there are many methods can be implemented in the prerocessing stage. One of them is Kimuras method for slant correction that using chain code for character slant prediction. Those method consumes time and resource in its computation, meanwhile the accuracy is not improved significantly. Therefore, this paper propose to create handwritten character into graph with string representation based on structural approach. The purpose is to provide ability in improving recognition accuracy without relying on normalisation technique. The similarity distance between graphs measured using levenshtein distance. Experiment conducted to recognize handwritten upper-case letters and digits character images which taken from ETL-1 AIST databases. Levenshtein distance has an accuracy of 84.69% on digits and 67.01% on alphabet with 5% size of data for training and value 10 for string representation length. As a comparison, the Kimuras method are implemented for slant correction which results in a reduction of accuracy until 6%. Comparisons also made with some of previous work.

ImNER Indonesian medical named entity recognition

We propose a medical named entity recognition for medical question answering system with Indonesian language. The aim is to provide a good medical named entity grammar by only using the available language resource. Our strategy here is to build the features most often used for the recognition and classification of medical named entities. We organize them along two different axes: word-level and list features, document and corpus features. For the reason we built our own features to Indonesian medical named entities and used it as the feature of the available with SVM Software. By using 3000 sentences, the highest accuracy score achieved is about 90%.

Direct skeleton extraction using river-lake algorithm

About 200 skeletonization algorithms have been developed since 1950. It shows that there is a large variety of the problem which arises from the combination of the criteria used whether in its accuracy (size and shape) or speed. In this paper, a new idea to enrich skeletonization problem by reversing the approach is proposed. This new idea has an advantage in the algorithm speed. The proposed approach collects points which are assumed as the center point in vertical and horizontal set which every vertical point is connected to horizontal set through projection. The result of the skeleton extraction has single-pixel width which represents starting shape and is conducted in high speed process.

Integration of Self-adaptation Approach on Requirements Modeling

Self-adaptation approaches appear to respond to environmental complexity and uncertainty of today’s software systems. However, in order to prepare the system with the capability of self-adaptation requires a specific strategy, including when conducting stage requirements modeling. Activity of requirements modeling to be very decisive, when selecting and entering new elements to be added. Here we adopt a feedback loop as a strategy of self-adaptation, which is integrated into a goal-based approach as an approach to requirements. This paper discusses the integration of the two approaches, with the aim of obtaining a new model, which has the advantages of both.

Shape based recognition using freeman chain code and modified Needleman-Wunsch

Contours are one of the most commonly used shape descriptors in object recognition problem. In this paper, we proposed object recognition system based on shape. The shape is obtained by extracting the contour of the object in the image using common techniques in image processing domain. Further, the shape is represented by using chain coding technique and the chain coded representation is modified into the set of segments, with each segment has a particular weight in accordance with its length in its polygonal approximation of the object shape. For the purpose of similarity calculation, we modified a common algorithm used in Bioinformatics field, namely Needleman-Wunsch algorithm, in the term of scoring function. We created a new definition and implementation of the substitution matrix (for the purpose of scoring function), according to the characteristics of set of line segment. From the experiment we have conducted, we successfully shown that the weight of each segment of the object shape has positive impact in the similarity calculation, shown by the precision and recall value.