Anita Thakur

CBCT image noise reduction and enhancement using Bi-Histogram method with bent activation function

Dental, oral and maxillofacial region are very sensitive so low radiation imaging technique like Cone Beam Computed Tomography (CBCT) is useful for investigation of diseases. But due to low radiation, quality of images is not fine they have noise and low contrast in visual impression. So for dentistry diagnostic purpose it is important to pre-process the CBCT image. This paper focuses on improvement of contrast, brightness and noise reduction of dental CBCT images by using Bi Histogram Equalization technique with bent activation function and adaptive median filter for noise reduction. The output image has been analyzed and evaluated using peak value ratio in term of Signal versus Noise (PSNR) and matrix of Structure in term of Similarity Index (SSIM). From the experimental results, it has been found that the adaptive median and Bi histogram with bent activation function has a better noise and enhancement capability as compare to convectional histogram method.

Autonomous car system using facial recognition and geo location services

Present day statistics of stolen vehicles used for crimes and accidents due to distracted driving or over speeding is alarming. Other frequent issues in todays commuting are over speeding and distracted driving. This paper aims to introduce an efficient method to reduce this issue in an advanced and novel way. The paper incorporates image processing and embedded system to realize this objective. Implementation of this system in automobiles prevents unauthorized use of vehicles and over speeding in speed limited areas. It can be easily implemented on the vehicles with a small modification; the cost of the device is very cheap compared to the other safety devices provided by different companies. Implementation of this technology on the vehicle will not affect the privacy and comfort of the traveler.

Analysis and Parameter Estimation of Microstrip Circular Patch Antennas Using Artificial Neural Networks

Microstrip circular patch antenna play a vital role in the mobile communication area. In round shape patch antenna, radius of round patch is important for generating the required resonant frequency. So proper functioning and high accuracy is required for design specifications of microstrip antenna. This paper discusses the design of round shape microstrip patch antenna using ANN. The mathematically calculated results using ANN model are compared with the Computer Simulation Technology (CST) to verify the results. The calculated resonant frequency is retained at S-band (2–4 GHz). S-band is used for RF power and high data rate application.

Metal Artefact Reduction from Dental CBCT Image Using Morphology and Fuzzy Logic

Cone beam computed tomography (CBCT) is a new-generation 3D image modality which is highly used in dentistry. As CBCT is a low-radiation imaging technique, reconstruction of image is prone to artefacts. Artefacts are the discrepancies between the original physical image and the mathematical modelling image process. In dental treatment, mostly metallic filling is done which produces metal artefact in imaging, in which it produces the reflection effect on imaging that misleads the diagnosis of treatment. In this paper, the proposed method reduces the reflection effect of metal artefacts and enhances the contrast of CBCT image. Here the proposed technique used morphological approach for reflection reduction, and fuzzy enhancement is used for contrast improvement. The output image has been analysed and evaluated using structure of similarity index matrix (SSIM) and peak value ratio in terms of signal versus noise (PSNR). Visual perception also shows the performance of the proposed work.

Black hole attack detection and prevention strategy in DYMO for MANET

Dynamic MANET On-Demand (DYMO) routing protocol has been used to establish an ad-hoc networks. DYMO is advance version of AODV routing protocol develop to improve the network performance. Security is the major challenge in DYMO routing protocol and prone for various security threats. This research work attempts to develop a mitigation algorithm to avoid and prevent genuine nodes from malicious attack. Black hole attack is one of the security threat in which the traffic is redirected to such a node that actually does not exist in the network. The black hole node presents itself in such a way to the other nodes and networks that it knows the shortest path. The complete research work is classified into three categories which are without attack, with attack and preventive scenario. The performance parameter taken for analysis are throughput and packet delivery ratio against the varying parameters like number of nodes, speed, pause time and area to observe the impact of black hole attack and proposed mechanism with different situation. A Qualnet 5.2 simulator has been used to simulate and evaluate the performance of proposed solution. The complete experimental setup concludes that improvement in mobile node increase the network performance but also increase the black hole impact. Subsequently, improvement in node speed degrades the black hole impact.

Contrast enchantment of CBCT images using Bi-Histogram method with soft plus activation function for dentistry

The Cone Beam Computed Tomography (CBCT) imaging technique is widely used to visualise dental, oral and maxillofacial region. CBCT imaging techniques are low radiation exposure on the patients. Due do this it produces low contrast images. Therefore, there is a need to enhance the low contrast CBCT image to highlight the important information that may aid diagnostic assessment, suitable treatment for dentistry. This paper focuses on improvement of contrast and brightness of dental CBCT images by using Bi Histogram Equalization technique with transform creation using activation function. The output image has been analysed and evaluated using Peak Signal-to-Noise ratio (PSNR) and Structure Similarity Index (SSIM). From the experimental results, it has been found that the soft plus activation function has a better enhancement capability.

Symptom & risk factor based diagnosis of Gum diseases using neural network

Gum diseases have multiple symptoms and risk factors are evolving. So Gum diseases are complex because multiple symptoms & risk factor affect the tooth loss in adults. Early prediction of gum disease helps to cure the tooth loss problem. Artificial Intelligent techniques are indeed worth exploring for diagnosis & prediction of complex medical problems. The aim of this paper is to determine a noninvasive method to detect whether a patient is suffered from gum disease. The problem addressed in this paper is prediction of possibility of periodontal and gingivitis gum disease and it is estimated from symptoms and risk factors given by the patients. A feed forward neural network with back propagation was used for prediction of gum disease.

Exploration of GBP2MP Network Performance for Next Generation Using Artificial Neural Network (ANN)

New era of world needed fast communication network for that optical fiber communication is promising solution. Optical networks are used in closed systems to open systems for various application like video on demand, voice over internet, video conference and real time broadcast. So fast performance criteria prediction of optical fiber network is time and cost saving solution. The aim of this paper is to determine the performance characteristic of Gigabit point to multipoint (GBP2MP) optical fiber network using artificial neural network. In artificial neural model (ANN), the input is frequency and fiber length in kilometres. Performance of optical network checked in term of minimum bit error rate (BER) parameters and results are discussed the performance of optical fiber network (OFN) when varying the length of fiber and frequency respectively.

Fuzzy contrast mapping for image enhancement

Human visual system appeases by a good contrast Images. Image enhancement techniques are best solution for improving the visual appearance of images to a human viewer. It also preserves the structure features of the image. Enhancement of the noisy image data without losing any significant information is very challenging. There are many uncertainties involved while capturing image and the performance of image enhancement varies with subject. It is well established that Fuzzy logic and fuzzy sets are very good at handling many uncertainties. The application of fuzzy theory to improve contrast of low contrast image is area of recent interest. To investigate and establish the application of fuzzy theory for enhancing a low contrast image, we propose a fuzzy based contrast enhancement of gray level images. Our experiment asserts that the proposed method has better performance than conventional methods. The image quality of the system has been evaluated based on visual appearance, peak signal to noise ratio and entropy.