Improved determination of the optimum maturity of maize based on Alexnet

Abstract

The increase in the number of humans and animals, particularly livestock in Sub-Sahara Africa without a correspondent increase in land resources has led to shortages, and consequently metamorphose into unhealthy clashes between farmers and herders. The unpredictable changes in climatic conditions in recent times and human activities has also contributed to deforestation and desertification. The maize plant is being considered to mitigate for the shortage by the application of Computational Intelligence technique and image processing in the determination of the optimum maturity of the maize. There are different varieties of maize that are quite suitable for different climatic conditions in Sub-Saharan Africa. In this paper, the optimum maturity of SAMMAZ 17 variety of maize seedling is selected due to its high resilient to drought, striga condition and its good composition of nutrients. The maturity is determined by the application of Alexnet on 3000 samples of maize comb captured at different maturity stages cultivated in the same farm land. The network gave an accuracy of 72.44%. The result obtained showed a 4.44% improvement over an earlier result obtained by the use of Resnet-50. The finding is a window of opportunity for improvement in the determination of the optimum maturity of maize.