Volume 13, Number 1

A vital step in automation of plant root disease diagnosis is to extract root region from the input images in an automatic and consistent manner. However, performance of segmentation algorithm over root images directly depends on the quality of input images.  During acquisition, the captured root images are distorted by numerous external factors like lighting conditions, dust and so on.  Hence it is essential to incorporate an image enhancement algorithm as a pre-processing step in the plant root disease diagnosis module. Image quality can be improved either by manipulating the pixels through spatial or frequency domain. In spatial domain, images are directly manipulated using their pixel values and alternatively in frequency domain, images are indirectly manipulated using transformations. Spatial based enhancement methods are considered as favourable approach for real time root images as it is simple and easy to understand with low computational complexity. In this study, real time banana root images were enhanced by attempting with different spatial based image enhancement techniques. Different classical point processing methods (contrast stretching, logarithmic transformation, power law transformation, histogram equalization, adaptive histogram equalization and histogram matching) and fuzzy based enhancement methods using fuzzy intensification operator and fuzzy if-then rule based methods were tried to enhance the banana root images. Quality of the enhanced root images obtained through different classical point processing and fuzzy based methods were measured using no-reference image quality metrics, entropy and blind image quality index. Hence, this study concludes that fuzzy based method could be deployed as a suitable image enhancement algorithm while devising the image processing modules for banana root disease diagnosis.

Hide Abstract

Cloud powered plant image warehouse serves as an instrumental solution for various scientific and academic personnel involved in research, education and extension at NARES (National Agricultural Research and Education System), through providing a common image dataset that complements for efficient and more productive activities by saving them time, space, hassle and financial resources. This web based image repository enables the entire scientific community to access the freely available resources contributed by the fellow researchers who worked on common areas of interest, besides facilitating to acknowledge the one who originally contributed. This also enables them to have better control and use of meta data with tagging and custom theme usage.

The Plant image warehouse has been developed by using XAMPP an open source platform, which works on the Cent OS, using Apache Web server and MySQL a relational web based data management system and PHP, the object oriented scripting language. The third party software used in developing this image warehousing database is ZenPHOTO, a configurable software system wherein the users are able to upload, search and share the images. The graphical user interface is restricted to static webpages where, upon request from the user, server sends the response unchanged, unless modified by the uploader. This potential plant image warehousing technique will outstand as an authentic and reliable source of plant image database to the entire working community at NARES (National Agricultural Research and Education System).

Hide Abstract

Deep learning is a very dynamic area in Sentiment Classification. Text analytics is the process of understanding text and making actionable decisions and acting on it. be it Amazon Alexa, Siri, Cortana everything is made up of Natural Language Processing. Text to speech and Speech to text are generating so many data sets every day. The internet has the largest repository of data, it is hard to define what to exactly do with it. sentiment are the opinions or the way of feelings of the public usually in the sequential form, in which many people face difficulty in living their daily life. Some are even ending their life just they are depressed. The approach here is to help the people suffering from depression with appropriate methodology to use in this work. Depressika: Early Risk of Depression Detection with opinions is a web application which detects the early risk of depression from the social media posts created by the users with appropriate Recurrent Neural Networks [RNN]. This is a classification problem of the Machine Learning [ML]. Depressika builds on Waterfall Methodology of application development using the Keras, Tensor  Flow, Scikit-Learn and Matplotlib to carryout and process sequential data and the overall process of development is carried out by Python programming Language.

Hide Abstract

The sustainability and environmental considerations have slowly become divergences, but having greatest influence in the supply chain management that must be contemplates to examine the environmental and organizational factors. The research considers environmental and sustainable strategies within companies, the efficient supply chain management strategies for manufacturers and consumers, and to the environment friendly product design and services, taking a case-by-case perspective and concentrating on enterprise businesses scale. Our finding reveals that green supply chain management firms are delivering exuberant environmental efficiency at an added cost. Among the identified obstacles we identified different obstacles and conceptual relations and barriers are graded based on dependency and driving sand. In future, green policies have greater customer services avenues thereby, appeal for suppliers, manufacturers and officials.

Hide Abstract

Cybersecurity systems are required at the application, network, host, and data levels. The research is purposed to evaluate Artificial Intelligence paradigms for use in network detection and prevention systems. This is purposed to develop a Cybersecurity system that uses artificial intelligence paradigms and can handle a high degree of complexity. The Pragmatism paradigm is elaborately associated with the Mixed Method Research (MMR), and is the research philosophy used in this research. Pragmatism recognizes the full rationale of the congruence between knowledge and action. The Pragmatic paradigm advocates a relational epistemology, a non-singular reality ontology, a mixed methods methodology, and a value-laden axiology. A qualitative approach where Focus Group discussions were held was used. The Artificial Intelligence paradigms evaluated include machine learning methods, autonomous robotic vehicle, artificial neural networks, and fuzzy logic. A discussion was held on the performance of Support Vector Machines, Artificial Neural Network, K-Nearest Neighbour, Naive-Bayes and Decision Tree Algorithms.

Hide Abstract