A new method for Tangerine tree flower recognition

Abstract

Different machine vision strategies are adapted for performing automated real time agricultural tasks in order to increase more productivity with less cost. Based on this notion, a new method is developed and implemented for detecting white color flowers in Tangerine tree and counting Tangerine fruit flowers to yield better outputs with regard to the existing schemes. Gaussian filter is employed to reduce unwanted noise in Tangerine tree flower recognition for Tangerine yield mapping system. It is observed that the newly developed method gives better valid output for tangerine tree flower detection in natural outdoor lighting, with different lighting condition without any alternative lighting source to control the luminance. The simulation result reveals that the new method is reliable, feasible and efficient compared to other existing methods. © 2012 Springer-Verlag.