Dissertations - M Tech (CS)

Permanent URI for this collectionhttp://164.52.219.250:4000/handle/10263/2147

These Dissertations were submitted in partial fulfilment of the requirements for the award of M TECH (Computer Science) Degree of Indian Statistical Institute

Browse

Subject: search.filters.subject.Convolutional neural network

Search Results

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Automated Determination of Glacier Ablation Zones
    (Indian Statistical Institute, Kolkata, 2024-06) Mondal, Anupam
    In recent decades, global temperature rises have significantly influenced glacier dynamics [1][2], underscoring the vital need for accurately delineating glacier boundaries to comprehend these shifts and document regional patterns. Despite this urgency, conventional methods struggle to map debris-covered glaciers (DCGs) due to their intricate nature. Climate change exacerbates glacier mass loss and intensifies glacier-related risks, necessitating ongoing monitoring and thorough analysis of climate-glacier interactions. Our research assesses the effectiveness of a convolutional neural network (CNN) in glacier mapping, utilizing Landsat satellite images, digital elevation models (DEMs), and DEM-derived land-surface parameters. Specifically, we seek to enhance the GlacierNet methodology by employing a CNN segmentation model to precisely identify regional DCG ablation zones. By training the models with satellite data from USGS and snow labeling from QGIS, and testing them on glaciers in the Karakoram region, we achieve improved estimations of the ablation zone, yielding high intersection over union scores. This study advances glacier mapping techniques, offering critical insights into climate change impacts on glacier dynamics in the Karakoram region. Furthermore, it marks a significant stride towards automating comprehensive glacier mapping, with potential applications in accurate glacier modeling and mass-balance analysis.
  • No Thumbnail Available
    Item
    Coke characterization: Segmentation of pores and constituents from microscopic images of Coke
    (Indian Statistical Institute, Kolkata, 2022-07) Biswas, Luna
    Coke is mainly used in steel industry as a fuel and a reducing agent for melting iron in the blast furnace, since it generates intense heat but little smoke. The quality of the coke material (like porosity, wall thickness, texture etc., as seen in a microscopic image of coke) affects the performance of blast furnace impacting the profit/loss of the industry. Therefore it is important to determine the structure and porosity of coke on a large scale. Manual process of coke characterisation is costly and slow. Automation of coke characterization, from microscopic images of cokes, is beneficial for the steel industry. An attempt has been made to calculate porosity of coke from the images, and produce semantic segmentation of the coke images into different types of metallurgical textures like inert, incipient, circular, lenticular etc. A shallow convolutional neural network (CNN) was trained with annotated coke images using cross entropy loss (between the probability distributions of the predictions out of the CNN and the target as per annotation, for different classes). A new contrastive loss function has been written, that maximises entropy between the probability distribution of a training sample with another sample belonging to a different class, in addition to minimising entropy loss between the probability distributions of the predictions and the target. This new loss function enables faster learning, and useful when quantity of annotations for training a model, is less. A shallow CNN model obtained higher accuracy in prediction of class for each pixel of the coke images, and the granularity of semantic segmentation was reduced when trained using this novel loss function.