Papers by Prema Kirubakaran
Knowledge Computing and its Applications, 2018
A pipeline crack is a major hazard to any type of liquid transportation. Oil industry depends mai... more A pipeline crack is a major hazard to any type of liquid transportation. Oil industry depends mainly on human detection of these cracks, which leads to many problems from health to environment disaster. To detect a crack, pipelines' inner layers are X-rayed, and these X-rays were later manually evaluated for cracks and holes. This technique evolves lot of time and resources. This proposed research work helps to diminish this problem by analyzing the cracks and holes through a computerized solution. A pipeline with crack is analyzed using image analysis and processing which comprises of various pattern recognition techniques. Image analysis and processing is one of the most powerful innovations in today's world. It brings all kinds of pattern recognition together and solves the problem of data identification and misuse of data. This is achieved by applying the method of pattern analysis and recognition. As a result, this technique of image analysis and processing is used to detect the holes and cracks which occur in a pipeline that carries any type of liquids and gases. This paradigm helps in environmental safety. As in the pipeline industry there are many man-made equipments and methods, computer application to carry out these process is lacking. To make a shift over to the computerized image recognition, high-frequency filter (HFF) with Gate Turn off thyristor (GTO) using unsupervised-based learning algorithm is implemented. Mathematical morphological operator and edge detection principles are used for image evaluation. Initially a digital camera with fiber optic cable is passed inside a pipeline to capture the cracked images. These images are converted as raster images and stored as bits. Later, these images are processed to view for hidden points using unsupervised cluster algorithm; after evaluating the hidden points, the crack has to be measured for its length and to identify the location where it occurs and this is achieved by developing mathematical morphological operator. Images are always
International Journal of Web Technology, 2016
Lecture notes on data engineering and communications technologies, Sep 22, 2022
IOSR Journal of Computer Engineering, 2014
This paper uses the technique of edge detection to make the image of an oil pipeline toappear bri... more This paper uses the technique of edge detection to make the image of an oil pipeline toappear bright and clear for further investigations after applying Mathematical morphology for curve error detection. This algorithm is developed to read the pixel values and eliminate the shadow images that appear. The pixels are studied carefully to give a set of appropriate values, which helps in setting up the right image for image evaluation. To achieve this, edge detection method is used and an algorithm is developed for image processing. This algorithm helps to build up a set of accurate pixel values, to determine the co-ordinates of the image for processing and analysis.This algorithm is developed mainly to remove the shadow images if caused during the camera intervention inside the oil pipeline.
A pipeline crack is a major hazard to any type of liquid transportation. Oil industry depends mai... more A pipeline crack is a major hazard to any type of liquid transportation. Oil industry depends mainly on human detection of these cracks, which leads to many problems from health to environment disaster. To detect a crack, pipelines’ inner layers are X-rayed, and these X-rays were later manually evaluated for cracks and holes. This technique evolves lot of time and resources. This proposed research work helps to diminish this problem by analyzing the cracks and holes through a computerized solution. A pipeline with crack is analyzed using image analysis and processing which comprises of various pattern recognition techniques. Image analysis and processing is one of the most powerful innovations in today’s world. It brings all kinds of pattern recognition together and solves the problem of data identification and misuse of data. This is achieved by applying the method of pattern analysis and recognition. As a result, this technique of image analysis and processing is used to detect th...
Materials Today: Proceedings, 2021
For the machine learning system, the selection of features is required because irrelevant attribu... more For the machine learning system, the selection of features is required because irrelevant attributes often affect the classification efficiency of this classifier. Selection of the features increases classification precision and decreases execution time for the model. For aortic and mitral setting models, equivalent to fifty predictions, the recurrent deep neural network (RDNN) was used. In several common medical diagnoses, hybrid machine learning approaches are used. This paper discusses numerous prediction and heart disease algorithms and compares genetic works along with neural networks.
Indian Journal of Science and Technology
International Journal of Advanced Engineering Research and Science, 2016
In the field of image analysis and processing, the post section of having a data record plays a v... more In the field of image analysis and processing, the post section of having a data record plays a vital role. The research work carried out for analyzing a crack image in an oil pipeline titled "Image Analysis and processing using mathematical morphological operators and high frequency filter for pipeline crack measurement" had a difficult phase of saving the data for future study. To overcome this issue a technique to preserve the image data is handled with the concept of big data analysis. A pipeline scanned for quality maintenance sends numerous pictures, where the pixel data is converted to binary data and then the calculated using the mathematical morphological operator based on erosion and corrosion of the images. These data are saved for further reference where the normal method of data saving using any hardware device was a big threat for loss of data and the cot to maintain the system is too high. To overcome this issue image data is been tried to save through the big data technique.
International Journal of Advanced Engineering Research and Science, 2016
In the field of image analysis and processing, the post section of having a data record plays a v... more In the field of image analysis and processing, the post section of having a data record plays a vital role. The research work carried out for analyzing a crack image in an oil pipeline titled "Image Analysis and processing using mathematical morphological operators and high frequency filter for pipeline crack measurement" had a difficult phase of saving the data for future study. To overcome this issue a technique to preserve the image data is handled with the concept of big data analysis. A pipeline scanned for quality maintenance sends numerous pictures, where the pixel data is converted to binary data and then the calculated using the mathematical morphological operator based on erosion and corrosion of the images. These data are saved for further reference where the normal method of data saving using any hardware device was a big threat for loss of data and the cot to maintain the system is too high. To overcome this issue image data is been tried to save through the big data technique.
This paper uses the technique of edge detection to make the image of an oil pipeline toappear bri... more This paper uses the technique of edge detection to make the image of an oil pipeline toappear bright and clear for further investigations after applying Mathematical morphology for curve error detection. This algorithm is developed to read the pixel values and eliminate the shadow images that appear. The pixels are studied carefully to give a set of appropriate values, which helps in setting up the right image for image evaluation. To achieve this, edge detection method is used and an algorithm is developed for image processing. This algorithm helps to build up a set of accurate pixel values, to determine the co-ordinates of the image for processing and analysis.This algorithm is developed mainly to remove the shadow images if caused during the camera intervention inside the oil pipeline.
Uploads
Papers by Prema Kirubakaran