®®®® SIIA Público

Título del libro: Proceedings - 2018 International Conference On Applied Mathematics And Computational Science, Icamcs.Net 2018
Título del capítulo: Automated Recognition of Mexican Bean Beetles Applying the LIRA Neural Classifier

Autores UNAM:
KAREN LUCERO ROLDAN SERRATO; TETYANA BAYDYK; ERNST KUSSUL; GRACIELA VELASCO HERRERA;
Autores externos:

Idioma:
Inglés
Año de publicación:
2018
Palabras clave:

automated detection; defoliating pest; Mexican Bean Beetle; LIRA neural classifier


Resumen:

The application of systems with bio-inspired intelligence developed for the monitoring and detection of environmental conditions and the management of pest diseases has become a high priority. Modern agricultural machines have begun to incorporate many electronic devices, intelligence software, and automatic and supervised processes to improve the quantity and quality of production in the field. These are key elements for the survival of humanity and the food demands of a growing population. In Mexico, the bean crop is one of the main crops for national consumption. The crop is severely attacked by the defoliating pests Mexican Bean Beetles (MBBs), which are the main cause of crop losses along with the excessive use of pesticides in crops. The goal of this work was to present an analysis of the results of automatic pest detection using neural networks. MBB recognition using an image database is based on the Greyscale Limited Receptive Area (LIRA) neural classifier. It was programmed and tested (C++). The image database contains 200 samples. The best recognition efficiency result was 85%.


Entidades citadas de la UNAM: