• M. Rajeshkumar
• Dr.S. Jayasankari

Smart farming methodologies are highly necessitated for growing or supporting crop yield to uphold emergent worldwide population with minimal environmental effect. Crop health spatial survey key indicators as well as treatment, e.g., herbicides, pesticides, besides fertilizers, merely to pertinent areas are attained through precision agriculture methods. In a similar way pertaining to crop health besides yield, selective weed treatment is regarded as crucial phase. Weed detection with least damage to surrounding plants to neighboring plants in consistent as well as precise way pose another challenge. Quad Histogram with Modified Convolutional Neural Network (MCNN) is utilized previously for categorizing crop as well as weed. Here this research exploits quad tree decomposition for feature extraction in addition Modified Convolutional Neural Network (MCNN) for classification. Still, color features are exploited for weed categorization. Classification rate is enhanced through several texture features which are also necessitated. In addition, single classifier is also sufficient to yield improved outcomes, henceforth ensemble learning is desired for prediction performance upgrading. An Ensemble Modified Convolutional Neural Network (EMCNN) is a promising solution for categorizing crop as well as weed merely. Basically, the input images in this research are Near-Infrared (NIR) and red images. Normalized Difference Vegetation Index (NDVI) from NIR as well as red patch images extraction is accomplished through simple automated image processing techniques. Noise removal from images in a proficient way can be attained through dynamically weighted median filtering algorithm which is followed by color feature extraction via Quad Histogram. Texture features removal is performed through Improved Local Binary Pattern (ILBP) along with shape parameters computation such as contour and skeleton features for performance enhancement. Dynamic Non-linear Decreasing Strategy based Glowworm Swarm Optimization (DNDSGSO) algorithm is greatly utilized for optimal features selection. Ensemble Modified Convolutional Neural Network (EMCNN) plays its role for sample categorization into crop, weed and background. Improved performance is achieved with the help of suggested system in contradiction to prevailing research pertaining to precision, recall, f-measure and detection rate which is validated through experimentation outcomes.

Normalized Difference Vegetation Index (NDVI), Improved Local Binary Pattern (ILBP), Crop and weed

"AN EFFICIENT CROP AND WEED CLASSIFICATION USING ENSEMBLE MODIFIED CONVOLUTIONAL NEURAL NETWORK ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 11, page no.897-915, November-2020, Available :http://www.jetir.org/papers/JETIR2011400.pdf

"AN EFFICIENT CROP AND WEED CLASSIFICATION USING ENSEMBLE MODIFIED CONVOLUTIONAL NEURAL NETWORK ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 11, page no. pp897-915, November-2020, Available at : http://www.jetir.org/papers/JETIR2011400.pdf