| Field cash crops occupy an important position in China, research on their characters is crucial in the process of mechanized production and harvest, which directly affect the production efficiency and product quality. The introduction of digital image technology not only brings efficient solutions to their characters research, but also shows broad application prospects. This paper analyzes the continuously improved image segmentation method, and divides it into traditional segmentation methods and image segmentation algorithm combined with specific tools. It mainly introduces the two popular methods, namely, segmentation method based on threshold and segmentation method based on edge detection, and wavelet analysis transform, genetic algorithm, active contour model, clustering algorithm and deep learning method combined with combined with specific tools. The image segmentation methods used in various algorithms are summarized, some applications in the characters of rape, cotton, soybean, and peanut are listed, and corresponding technicala framework diagram added in the genetic algorithm, active contour model, clustering algorithm and Fast R-CNN algorithm in deep learning. Finally, the main problems of the application of image segmentation technology in field cash crops are analyzed, and the prospects are put forward. |
