Mobile glass is a vital part of smart phone which is subjective to frequent breakage. The fragile glass is a fundamental part of various items that are utilized each day, frequently without visualizing it. Plainly present-day life would not be conceivable without glass materials. E-commerce organizations spend a lot of man power and time in the exchange process of mobile phones both with the customers and the manufacturers. They need to check the condition of the mobile screen glass display during an exchange process with their customers ordering, re-selling process to the manufacturers. The smart phone in ideal condition can be used by e-commerce organization to resell the it under a deal, showing the phone as ‘Used/Refurbished’. The malfunctioning smart phone can be thrashed based on parts of it. Some parts can be used for research and development process. Hence there is need for automated framework for above said process. The convolutional neural network is an efficient process for implementation of it. The proposed system focuses on designing a glass classification model based on the level of damage of the glass for prediction of the damaged mobile phones into respective category. This problem falls under multi-class classification problem for detection of glass screen damage level with predefined models like Densenet, Resnet, Squeezenet and custom models. The real-time images are collected for training the model and validated with the sample images for improved accuracy. The proposed model shows an accuracy of 85% and performance can be further improved with the training with large variants of augmentation techniques. The model can be generalized with the variants of glass images and further finds its scope in variety of applications.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Glass Damage Classification in Mobile Phones using Deep Learning Techniques


    Contributors:


    Publication date :

    2021-12-02


    Size :

    1080923 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Expression of Concern for: Glass Damage Classification in Mobile Phones using Deep Learning Techniques

    Selvi, K Tamil / Thamilselvan, R / Pratheksha, K et al. | IEEE | 2021

    Free access

    Classification of Damage to Car Tires Using Mobile-Based Deep Learning

    Sukmana, Septian Enggar / Rahmad, Cahya / Khairy, Muh. Shulhan et al. | IEEE | 2024


    Safe Driving Using Mobile Phones

    Fazeen, Mohamed / Gozick, Brandon / Dantu, Ram et al. | IEEE | 2012


    Mobile phones advance

    Tema Archive | 1978


    Road Safety Implications of Using Mobile Phones while Mobile

    Mittal, N. / Chakrabarty, N. / Sarin, S. M. | British Library Online Contents | 2005