Informatics Publishing Limited

S. Sandhu Mandeep ¹, K. Kamra Amit ², Vijay Dhir

Affiliations
1 Department of CSE, BMSCE, Muktsar, IN
2 GNDEC, Ludhiana, IN

Abstract

A novel approach to detect microcalcification in the mammographic images is presented in this paper. Microcalcification are small tiny spots of calcium having varying size and shapes spread in various part of the breast. The proposed approach is useful to enhance these regions to locate the regions having this type of cancer. The approach makes use of Discrete Wavelet Transform to decompose the image into detailed coefficient. Then these spatial decomposed subbands are subjected a dynamic thresholding technique to enhance the region. This technique eliminates the need to perform the global thresholding. This becomes a pure computer based approach to find malignant and benign microcalcifications.

Keywords

DWT, Benign, CAD, Malignant, Thresholding.

Full Text

     

Dinesh Kumar ¹, Vijay Dhir ², Rakesh Kumar ³

Affiliations
1 Department of Information Technology, DAV Institute of Engineering & Technology, Jalandhar-144008, Punjab, IN
2 Department of CSE/IT, Sant Baba Bhag Singh Institute of Engineering Technology, Padhiana, Jalandhar, Punjab, IN
3 Department of CSE/IT, Sant Baba Bhag Singh Institute of Engineering & Technology, Padhiana, Jalandhar, Punjab, IN

Abstract

The purpose of this research is to test the efficiency of wavelet based noise removal and compression techniques to remove noise and compress the signals using MATLAB. In the present investigation, the signals of interest are normal images but intended to extend the current work to process MRI images. Various kinds of wavelets have been considered for this purpose like Daubechies: 'db1' or 'haar', 'db2', ... , 'db45', Coiflets: 'coif1', ... , 'coif5', Symlets : 'sym2', ... , 'sym8', ... , 'sym45', Discrete Meyer wavelet: 'dmey'.

Keywords

Threshold, Entropy, Compression, Wavelet, Denoising.

Full Text

     

Vijay Dhir ¹, Amit Kamra ², Rakesh Kumar ¹, Baljeet Saini ¹

Affiliations
1 Padhiana, Distt: Jalandhar, Punjab, IN
2 GNDEC, Ludhiana, Punjab, IN

Abstract

A number of video clips are available online. Users are uploading videos and provides the source of indexing information as title of video and set of keywords. Automated extraction of video content from a large scale video is a challenging and yet unsolved problem. Proposed method finds the audiovisual mapping. All pieces of information is trained automatically without any human supervision. We presents the results in 1200 videos and show the effectiveness of the method per celebrity basis.

Keywords

Speaker Recognition, Face Recognition, Diagonal Covariances, Equal Error Rates.

Full Text

     

Ramandeep Kaur ¹, Amit Kamra ², Vijay Dhir ³

Affiliations
1 RBIEBT, Sahuran, Punjab, IN
2 GNE, Ludhiana, IN
3 Sant Baba Bhag Singh, Hoshiarpur, IN

Abstract

Fingerprint recognition is a method of biometric authentication that uses pattern recognition techniques[1] based on high-resolution fingerprints images of the individual. Fingerprints have been used in forensic as well as commercial applications for identification as well as verification. The fingerprint surface is made up of a system of ridges and valleys. The steps for Fingerprint recognition include image acquisition, preprocessing, feature extraction and matching[2]. In the present work, a new fingerprint feature detection algorithm has been proposed. It has been found that presence of noise in fingerprint images leads to spurious minutiae. To overcome this problem, feature extraction has been done which efficiently determine the minutiae points in fingerprint [1]. The proposed method can be used in matching the template for finding bifurcation and termination. The new smoothing algorithm is proposed for the detection of the features of fingerprints. A method has been introduced for finding ridges in the fingerprint image with the help of eight different masks. It is a process of making a binary image of ridges from the grayscale fingerprint image. The experimental results showed the accuracy of the algorithm in terms of genuine acceptance rate, false rejection rate, false acceptance rate.

Keywords

Bifurcation, Orientation Field, Ridges Segmentation and Smoothing, Termination.

Full Text

     

Lovely Sharma ¹, Vijay Dhir ², Kamaljeet Kaur ³

Affiliations
1 Punjab Technical University, Kapurthala Road, Jalandhar – 144601, Punjab, IN
2 St. Soldier Institute of Engineering and Technology, Jalandhar – 144622, Punjab, IN
3 Department of Computer Science, D. A.V. College, Amritsar – 143001, Punjab, IN

Abstract

Objective: Currently, the user is rely on the results of the Google search i.e. lengthy documents, web links etc. So, Question Answering (QA) dialogue system becomes more vital to access the succinct answers in a jiffy. Methods: The main focus is on Query Processing based on partial semantic and syntactic analysis and keyword based approach. The major module is Dialogue Manager which is responsible to communicate between the User and the System to resolve anaphora and Co-reference problems. Findings: The existing Indian Railway system is based on clicking system which is more time consuming. Thus, proposed QA dialogue system provides specific and concise answers to users. The results are encouraging and a motivation for further research in this field. The proposed system restricted to provide answers of related queries but further extension of the system can also include replies related arrival time, departure time over the phone etc.

Keywords

Dialogue based Question-Answering System, Dialogue Manager, Semantic Analysis

Full Text

   

Lovely Sharma ¹, Vijay Dhir ², Kamaljeet Kaur ³

Affiliations
1 Department of Computer Science and IT, Doaba College, Jalandhar - 144001, Punjab, IN
2 St. Soldier Institute of Engineering & Technology, Jalandhar - 144622, Punjab, IN
3 Department of Computer Science, D.A.V College, Amritsar - 143001, Punjab, IN

Abstract

Objectives: The plethora of information available on the web makes it difficult to search for a specific and suitable query. So, Question Answering dialogue system becomes helpful to provide precise and short answers to users instead of lengthy documents or passages. Method: The main focus is on keyword based approach and partial linguistic analysis to process the query. The main component is Dialogue Manager which clarifies the conversation between user and system to resolve the elliptical and anaphoric problems. Findings/ Application: The paper is based upon the development of Question answering dialogue system in Hindi language for the domain of Railways. The study is restricted to Northern India alone. The existing system of Indian railway does not support Question Answering system while this novel system will provide anytime anywhere facility of enquiry about railway. The implemented system is quite simple, interactive and easy to use such that end user can ask the query using Hindi transliteration. The results are quite satisfactory as precision and recall values are 98% and 97% respectively.

Keywords

Dialogue Manager, Partial Syntactic and Semantic Analysis, Query Processing, Question Answering Railway System.

Full Text

   

Harpreet Kaur ¹, Vijay Dhir ²

Affiliations
1 Department of Computer Science and Engineering, Punjab Technical University, Kapurthala, Punjab - 144601, IN
2 Sant Baba Bhag Singh Insitute of Engineering & Technology, Jalandhar, Punjab - 144020, IN

Abstract

Objectives: This paper presents a new feature extractor which we named as Local Color Oppugnant Mesh Extrema Patterns (LCOMeEP) for retrieval of images. Methods/Statistical Analysis: The suggested method gathers the colortexture statistics among the RGB (Red, Green and Blue) and gray scale of the given image. The color-texture statistics is extracted based on mesh extrema which assembles the relationship among the neighbors using extrema. Findings: The proposed method is diverse from the Directional Local Extrema Patterns (DLEP) collect the directional information which is based on local extrema in an image. Whereas the method presented in this paper, (LCOMeEP) gathers the mesh extremas amid the RgG (red, gray, green), GgB (green, gray, blue) and BgR (blue, gray, red) spaces. The enactment of the presented technique is amended by assimilating the LCOMeEPs with histograms of HSV (hue, saturation, value). The enactment of the research work is estimated by simulating on standard data sets, Corel-5K and Corel-10K in context of recall, precision, Average Retrieval Rate (ARR) and Average Retrieval Precision (ARP). Application/Improvements: The outcome after inspection illustrates a substantial enhancement as related to the contemporary features for image retrieval.

Keywords

Image Retrieval, Local Binary Patterns (LBP), Local Extrema Patterns, Local Oppugnant Patterns, Pattern Recognition, Texture

Full Text