AgroSearch

Autor: Sunil More, Mininath K. Nighot
Rok vydání: 2016
Předmět:
Zdroj: Proceedings of the Second International Conference on Information and Communication Technology for Competitive Strategies.
DOI: 10.1145/2905055.2905102
Popis: Uneven climate change demands improved and modern methods for agriculture domain. Also increased demand for food to accommodate global population, farmers try to multiply food production. Automation and intelligent decision system helps to accomplish this mission. AgroSearch is web based search tool help farmers to identify diseases and pests for pomegranate as well as provides remedy. User gives the query in the form of text, image and visual image click and gets results like search engine. It follows image preprocessing, feature extraction, training and testing respectively. Features extracted from images are color, color coherence vector and morphology. Clustering is done by applying K-Means algorithm. Finally classification of image using Multi-class Support Vector Machine into one of the class. Morphology feature gives better results. Accuracy of system is 84% and validated experimentally.
Databáze: OpenAIRE
Pro tento záznam nejsou dostupné žádné jednotky.