Insight Analysis of Deep Learning and a Conventional Standardized Evaluation System for Assessing Rice Crop's Susceptibility to Salt Stress during the Seedling Stage

• Published in: SN computer science Vol. 4; no. 3; p. 262
• Main Authors: Shiragudikar, Sharada K., Bharamagoudar, Geeta, Manohara, K. K., Malati, S. Y.
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.05.2023; Springer Nature B.V
• Subjects: Advances in Computational Intelligence for Artificial Intelligence; Artificial intelligence; Automation; Classification; Computer Imaging; Computer Science; Computer Systems Organization and Communication Networks; Crop yield; Data Structures and Information Theory; Deep learning; Evaluation; Information Systems and Communication Service; Internet of Things and Data Analytics; Machine Learning; Pattern Recognition and Graphics; Review Article; Rice; Salinity; Seeds; Software Engineering/Programming and Operating Systems; Vision; India; Deep learning; Salinity stress; Seedling stage
• ISSN: 2661-8907; 2662-995X; 2661-8907
• DOI: 10.1007/s42979-022-01656-2

One of the most used food crops produced worldwide is rice. However, there is a huge possibility that excessive saline levels, particularly at the seedling stage, may negatively impact rice productivity. Therefore, it is essential to identify and create salinity-tolerant rice crop varieties as soon as possible, especially at the seedling stage, to avoid a reduction in rice yield. For the classification of visual signs and the traditional method of standard evaluation system for identification of rice crop salinity stress, human expertise is needed. Most of the time, this can result in errors that in turn result in incorrect classification and also its time consuming. To identify and classify salinity stress in rice seedlings using field images, the study reports the need of deep learning built model over traditional method of assessing rice crop's susceptibility to salt stress during the seedling stage. So to build the classification model, we use the pre-trained VGG 16, one of the deep learning techniques, which is designed in Jupyter Notebook using Python programming. The model is able to classify the rice seedling images according to the scores as Grade1, Grade3, Grade5, Grade7, Grade9 that supports to the fact that there is a huge need of a computer-based classification system for the salinity prediction that could be used as tool for automating the classification process for rice development to assist scientists and agriculturalists in the rice crop management system.