Informatics Publishing Limited

R. Jayadurga ¹, R. Gunasundari ²

Affiliations
1 Department of Computer Science, Karpagam University, Coimbatore - 641021, Tamil Nadu, IN
2 Department of Information Technology, Karpagam University, Coimbatore - 641021, Tamil Nadu, IN

Abstract

Objectives: To increase the classification accuracy and the performance of the artificial neural network classifier based on hybrid feature extraction. Two different literatures are hybrid together to obtain the better result. Methods/Analysis: Identification and classification of vehicle object system are based on a hybrid feature extraction method and neural classifier. Every image is divided in two equal size 10x10 sub-block. From each sub-block of the image, central moment and geometrical moment features are extracted without pre-processing. The extracted feature vector is normalized and combined together. Normalization is done by using ZScore normalization technique. Then the normalized feature vectors are fed to the Artificial Neural Network (ANN) classifier by using Feed Forward Back Propagation Algorithm (FFBPA) for classifying the vehicle object. Findings: Illinois at Urbana-Champaign (UICI) standard database is used for vehicle object classification. UIUC Dataset contains 500 car images and 500 non-car images with mixed background. The normalized input feature vectors which have been selected are improving the classification accuracy compared with the previous work. It increases the true categorization ratio and decreases the false categorization ratio. The quantity improved performance result shows 95.3% compared with a similar work of various literature methods. Applications/Improvements: This novel method plays a vital role in applications such as vehicle security system, traffic monitoring system etc.

Keywords

Artificial Neural Network, Back Propagation Algorithm, Central Moment Features, Feature Extraction, Geometrical Moment Features, Hybrid Feature, Normalization, Vehicle Classification.

Full Text

   

R. Jayadurga ¹, R. Gunasundari ²

Affiliations
1 Department of Computer Science, Karpagam University, Coimbatore - 641 021, Tamil Nadu, IN
2 Department of Information Technology, Karpagam University, Coimbatore - 641 021, Tamil Nadu, IN

Abstract

Objectives: To increase the performance of the classifier for the vehicle object among a mixed and highly texture background using hybrid feature extraction method without pre-processing. Methods/Analysis: Vehicle Object recognition system performance is based on the hybrid of feature vector extraction method and artificial neural network classifier without pre-processing. Every image is divided into single blocks size with 20x20 each. The feature vector is extracted from each single size block of the picture. Normalization is done for the extracted feature vector of the vehicle object in the image. These feature vectors are given as input to the neural network classifier for classification. The feed forward Back Propagation Neural Network (BPNN) algorithm is used to train and test the input feature vector by using Neural Network Classifier (NNC) for the vehicle classification. Findings: The idea of the proposed method is that combining the two different literatures namely statistical and spectral texture features without pre-processing for classification. This method is trained and tested with Illinois at Urbana-Champaign (UIUC) standard databases. UIUC database contains car and non-car images with mixed and highly textured background. The findings indicate that the selected input feature vector improves the classification accuracy rate compared to the previous literature. Also the hybrid features maximize the correct classification and minimize the wrong classification. The improved performance results 90.1% of quantitative evaluation is compared with different literature methods of similar work. Applications/Improvements: In different applications, the proposed method plays vital part in surveillance, security for vehicles, monitoring the traffic, etc.

Keywords

Back Propagation Algorithm, Feature Extraction, Hybrid Feature, Neural Network Classifier, Normalization, Statistical Features, Spectral Texture Features, Vehicle Categorization.

Full Text

   

Mohammad Shahnawaz Shaikh ¹, R. Jayadurga ², Jagannath Jadhav ³, Shadab Ahmad ⁴

Affiliations
1 Department of Electronics and Telecommunication Engineering, GH Raisoni Institute of Engineering and Technology, IN
2 Department of Computer Science, Soundarya Institute of Management and Science, IN
3 Department of Electronic and Communication Engineering, K.L.E. College of Engineering and Technology, IN
4 Department of Electronics and Communication Engineering, Samara University, ET

Abstract

In general, an antenna is an interface that transmits signal data and receives incoming signal data. The radio waves received through this interface help to do the necessary things for the transmitter and receiver circuit systems used there. Also a radio transmitter antenna transmits different waves generated from the current generated at its tip to different areas. In this paper, the functions of increasing its bandwidth by making changes in some dimensions of the antenna are proposed. The oscillating current used in its transmitter area increases its vibration waves. This increases the amount of airwaves generated there and the number of data transmitted through it. So its bandwidth is more likely to be high. Furthermore these functions generate varying magnetic fields so that the time taken by the cross-sectional magnetic fields of the antenna varies.

Keywords

Antenna, Radio Waves, Electric Current, Radio Transmitter, Bandwidth, Broadcasting, Two-Way Radio

Full Text

     

R. Jayadurga ¹, M. Sathiya ², G.K. Arpana ³

Affiliations
1 Department of Computer Science, Soundarya Institute of Management and Science, India., IN
2 Department of Information Technology, Karpagam Institute of Technology, India., IN
3 Department of Electronics and Communication Engineering, East West College of Engineering, India., IN

Abstract

Most of the time, these classifiers are trained using general-purpose datasets with a lot of classes. Therefore, the performance of these classifiers may not be as good as it could be. Both choosing classifiers based on registrations and dividing them into groups based on the subjects they cover are possible solutions that could lead to better classifier performance. This makes it clear that a classifier division and selection strategy needs for the proposed optimization to work. With the help of this method, the proposed model for feature extraction can choose an appropriate classifier while taking subscription constraints into account. There are subscriptions with the best values of n, and the results of using only n-class classifiers from one domain and ignoring classes from other domains are also given. These are in the same place as the effects of only using n-class classifiers from a certain domain. In this article, these are talked about in the same context as what happens when you only use n-class classifiers from a certain domain. For high-performance use of SAE-based systems, you need to use a classifier selection technique. This method is also needed for the investigation of multimedia events that need the method. To establish the effectiveness of the multimedia event-based system as well as its dependability, we are making use of traditional evaluation methods such as throughput and accuracy. These measures include the following: When compared to the efficiency of the system when using a classifier with a single class, the efficiency of the system diminishes as the number of classes per classifier increases. This is the case regardless of the other measures. This is the situation about both the throughput and the precision of the operation.

Keywords

Multimedia Data, Stacked Auto Encoder, Deep Learning, Classifier.

Full Text