An Improved Image Processing Algorithm on Monitoring Different Growth Stages of Bamboo Shoots

Image processing to monitor growth height of bamboo shoots should consider color features due to the prominent color difference between bamboo shoots and background. Bamboo shoots can exhibit significant color contrast compared to mature bamboo, and also show noticeable color change process at various growth stages. Therefore we study different image processing method according to the corresponding growth stage, i.e. early stage and mid stage. It's found that separately using H-component of HSV model can produce more ideal segmentation result at the early stage, while only combining S- with H-components using a suitable mixing coefficient <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\boldsymbol{k}$</tex> can ideal segmentation result occur at the mid stage.