Research On Vision Based Measurement System For Plant Physiclogical Information

Plant physiological parameters such as stem diameter, area of leaf, leaf angle, position of the tip of leaf are of great importance to the research on plant physiology. The prevailing instrument of such parameters is based on the mechanical or physical sensor. Some researchers use the camera as a tool for such research, but without the specified instrument. The Computer Vision approach in this area will enable us to measure and quantize the information we haven't measure before such as the continually measurement of area of plant leaf or the position of the leaf. Computer Vision in such area has its advantage.The content of this thesis is as following:1 Measuring of the Stem Diameter: We use the embedded Smart Camera model VC4472 as the main component to build a Plant stem diameter measuring system.(including the optics mechanical) Developed the software system on the real-time operating system VC/RT 5.0 which has the function of image acquisition, measurement algorithm, data storage, network communication. This device is tested in a real Green House environment and acquired experimental data to verify the validity of the measurement system.2 Measuring the area of plant leaf: We use single camera to measure the leaf of planar type. Use the differential method to extract the leaf from its background. This method can regulate the influence of the nature light change. We use the texture information to locate the leaf in the coordinate of the camera which is a key problem in the single camera system to measuring plant leaf.3 Measuring the leaf angle and tip position. We use Camshift (Continuously Adaptive Mean-Shift) and Harris algorithm to this system. The hardware system is based on the PC and USB camera. We use VC++6.0 and OpenCV to build the software system for measuring such parameters. This software system can achieve the functions of image acquisition, automatic recording, data storage, etc.