| The growth structural parameter of crops stem root plays a very important role in the intelligent field operation, it can not only make the weeding machine independent avoidance crops, also can provide the basis for the crops root on precision fertilization injection nutrient and the branches pruning. Previous to extract the structural parameters of crop growth mainly rely on artificial directly measured, which has a larger workload and low efficiency, at the foundation of reading and summarized the domestic and foreign related research, this article regards Tomatoes as the research object, adopting the method of image processing to extract its structure parameters, which provide some necessary basis for further intelligent operations.The main research content of this article is as follows:(1)The field image acquisition method has been studied. In order to reduce the system cost choose USB camera as the core component of the visual system, at the same time adopt C language to achieve the development of USB camera driver based on the VC++6.0, and complete image acquisition and processing, which laid a subsequent foundation for the further research.(2)The image pretreatment and gradation treatment have been studied. Due to the collected images in the field would contain a lot of noise, if not try to make some pretreatment measures which would bring certain difficulty to the subsequent processing, during the image pretreatment by comparison, this article finally choose the method of 3x3 neighbor average filter for image pretreatment; considering the influence of light on image processing, we selected normalized 2.25g-r-b to image gradation after pretreatment.(3)The division of gray image has been studied. Through the comparison of different methods, finally adopted the Otsu automatic threshold method which has stronger adaptive ability to threshold the image, in order to facilitate subsequent processing, the divided image was made some morphology processing.(4) Extraction of tomato stem root growth structural parameter information has been studied. Through the comparative study among the projection method〠the stem root block secondary projection method〠the image rotation method and the stem root block image rotation method, finally choose the stem root block image rotation method to locating the crops stem root, during the process of image rotation, realize the extraction of crop growth angle, and on this basis, complete the extraction of the crop thick stems and stalks. |
