Plant counting with low altitude image procassing

Abstract

Image processing techniques with the development of the technology began to be used widely in the field of agriculture. Studies often benefit from satellite images. However, in recent years, because of time saving, low cost and less dependence on weather conditions, the unmanned aerial vehicle (UAV) images are also used. Because of their own characteristics, UAV records in studies on small fields and gardens, are more advantageous than satellite images. In this study, the low-altitude aircraft with low-resolution images were taken with the image processing approach aimed to plant counts. Results in line with objectives set out by applying various image processing techniques have been tried to get the picture given as input. Background subtraction, gray level image transformation, thresholding, black and white levels of the image conversion, etching processes are the main techniques used. Mentioned techniques in MATLAB has been implemented in environment. Working examples are tested for an input image obtained 89% success. © 2015 IEEE.