Research Cell: An International Journal of Engineering Sciences

Gurpreet Singh ¹, Jyoti Rani ¹

Affiliations
1 GZSPTU Campus, Bathinda, IN

Abstract

The "Adaptive Approach for Historical or Degraded Document Binarization" is that in which Libraries and Museums obtain in large gathering of ancient historical documents printed or handwritten in native languages. Typically, only a small group of people are allowed access to such collection, as the preservation of the material is of great concern. In recent years, libraries have begun to digitize historical document that are of interest to a wide range of people, with the goal of preserving the content and making the documents available via electronic media. But for historical documents suffering from degradation due to damaged background, stained paper, holes and other factors, the recognition results drop appreciably. These recognition results can be improved using binarization technique. Binarization technique can differentiate text from background. The simplest way to get an image binarized is to choose a threshold value, and organize all pixels with values greater than this threshold as white, and every other pixels as black. The problem arises, how to select the correct threshold. The selection of threshold is performed by two methods: Global, Local. Our main focus is to effectively binarize the document images suffering from strain&smear, uneven backround, holes&spot and various illumination effect by applying Adaptive Binarization Techniques. Our objectives is to Study various Traditional Binarization Techniques and to develop a hybrid binarization technique which will be more efficient than traditional techniques in term of noise suppression, text extraction and enhance the document to make it better for readability&automatic Document analysis. Result is analyzed and obtains which conclude that.

Keywords

Global, Local, Binarization, Illumination, Hybrid Binarization, Historical Documents.

Full Text

   

Gurpreet Singh ¹, Baljit Singh ²

Affiliations
1 IET, Bhaddal, Ropar, IN
2 BBSBEC, Fatehgarh Sahib, IN

Abstract

Computer has been widely deployed to our daily lives, but human computer interaction still lacks intuition. Researchers intend to resolve these shortcomings by augmenting traditional systems with human like interaction mechanism. Today, dedicated hardware often infers the emotional state from human body measures.These have been a considerable amount of research done into the detection and implicit communication channels, including facial expressions. Most studies have extracted facial features for some specific emotions in specific situations. In this paper we uses a feature point tracking technique applied to five facial image regions to capture basic emotions. The used database contains 219 images, 10 Japanese female, six expressions and one neutral. We use grayscale images which are ethically not diverse. We use optical flow based analysis to detect emotions from human facial image data. Our proof of data demonstrates the feasibility of our approach and shows promising for integration into various applications.

Full Text

   

Gurpreet Singh ¹, Raman Maini ²

Affiliations
1 Department of Computer, Punjabi University Patiala, Punjab, IN
2 Punjabi University, Patiala, Punjab, IN

Abstract

Mobile ad-hoc networks are known by their continuously changing topology, no permanent infrastructure, resource restrictions and multi-hop scenario. Owing to this, they are more exposed to security attacks. In these networks, one of the hazardous attacks is packet dropping attack differentiated into 2 types: 1. Black hole and 2. Gray hole. Within both of them, an attacker replies to source node incorrectly that it is encompassing the shortest path to the destination. In former, attackers drops the entire volume of packets sent through it and in latter, attacker drops packets selectively. The following study implement to detect and isolate these attacks in network. The performance of the network has been compared without black hole node and with black hole node on the basis of PDR, throughput and energy consumption of the network. It has been observed that the actual performance of the network without black hole node is better as compare with the network with black hole node.

Keywords

Mobile ad hoc Networks, Black Hole Attack, Grey Hole Attack, Throughput.

Full Text

   

Ramanpreet Kaur ¹, Gurpreet Singh ¹

Affiliations
1 Computer Engineering Department, Punjabi University, Patiala, IN

Abstract

Diabetes mellitus is a chronic disease and a major public health challenge worldwide. According to the International Diabetes Federation, there are currently 246 million diabetic people worldwide, and this number is expected to rise to 380 million by 2025. Diabetes is a standout amongst the most well -known non- transmittable diseases in the world .Vast amount of data available in health care industry is difficult to handle, hence mining is necessary to find the necessary pattern and relationship among the features available. Medical data mining is one major research area where evolutionary algorithms and clustering algorithms play a vital role . Several data mining and machine learning methods have been used for the diagnosis, prognosis, and management of diabetes. Several researchers are using statistical and data mining tools like rapid miner ,weka , KNIME etc. to help health care professionals in the diagnosis of diabetes. The data source for this research is taken from UCI repository. Various experiments are made iteratively by using various techniques on Tanagra and R tool. In this research work, K-Means is used for removing the noisy data and genetic algorithms for finding the optimal set of features with Support Vector Machine (SVM) as classifier for classification. It shows that the proposed method using Tanagra with an accuracy of 76.44 % has attained better results compared to R Tool with an accuracy of 75.39 %.

Full Text

   

Gurpreet Singh ¹, C. P. Kamboj ²

Affiliations
1 Research Scholar, Department of Computer Science, Punjabi University, Patiala, IN
2 Assistant Professor, Punjabi Computer Help Center, Punjabi University, Patiala, IN

Abstract

This paper provides a comprehensive review of the role of deep learning techniques in natural language processing (NLP). With the explosion of textual data in recent years, the need for efficient and accurate NLP algorithms has become increasingly important. Deep learning approaches, which are based on neural networks, have shown great potential in addressing many NLP tasks such as language modeling, sentiment analysis, text classification, and machine translation, among others. In this paper, we provide an overview of the key deep learning models that have been used in NLP, including recurrent neural networks (RNNs), convolutional neural networks (CNNs) and transformers. We also discuss the challenges and limitations associated with deep learning models, such as overfitting, data sparsity, and interpretability. Moreover, we review the recent advancements in deep learning techniques, including transfer learning and pre-training, which have enabled the development of state-of-the-art NLP models. Finally, we highlight some of the promising future directions in the field of deep learning for NLP, such as multi-task learning and the integration of symbolic reasoning with neural networks.

Keywords

Natural Language Processing, Deep Learning, Artificial Intelligence, Neural Networks, Language Modeling.

Full Text

   

Gurpreet Singh ¹, Dr. Sukesha Sharma ¹

Affiliations
1 UIET, Panjab University, Chandigarh, IN

Abstract

In this review paper an introduction to Parkinson disease, stages in Parkinson disease, causes and the types of Parkinson disease has been discussed. Wearable devices or sensors that can be used for detection of Parkinson disease, checking the accuracy of the results or data provided by the wearable/portable sensors with clinical recorded data has also been discussed. Classifiers that can be used in machine learning for the detection of Parkinson disease have also been discussed. Methods by which Parkinson disease can be detected by medical lab tests have also been discussed. Also, we have discussed the prediction model of Parkinson disease using CNN in machine learning. All results are finally tabulated for comparison.

Keywords

Parkinson Disease, Wearable Devices, Machine Learning, Deep Learning, Machine Learning Classifiers.

Full Text