Recognition and location of crop seedlings based on image processing

With the development of digital image technology, we can easily obtain a large number of crop growth images. Through effective analysis of the image, the growth information of crops can be obtained, which can better direct agricultural production. The efficiency of traditional seedling growth monitoring is low, especially in large-scale farmland, which takes a lot of time. Artificial method timely restricts scientific decision-making of cultivation crops. The progress of machine vision and image processing technology provides a new way for harmlessly monitoring of crop seedling growth .The results of image analysis can help agricultural producers to understand the growth of crop seedlings quickly and accurately, so as to take effective management as soon as possible. In this paper, the images of sunflower seedling collected in farmland environment are taken as the research object. The main research content is to segment green crops from soil background. Segmentation method of sunflower seedling image based on color features and Ostu threshold segmentation is proposed. The method is simple in calculation, and can adapt to the segmentation of farmland environment images, which lays the foundation for crop recognition process. Based on the image recognition results, the algorithm locates the seedlings. Through the rapid identification of sunflower seedlings, it is possible to fill the gaps with seedlings where the seedlings are less distributed. On the contrary, if the seedlings are too dense, the number of seedlings needs to be reduced. The algorithm provides a basis for precise management. The results show that the algorithm with extra green feature can quickly and effectively identify sunflower seedlings from background, and locate the seedlings based on the image recognition results. This algorithm is not sensitive to soil moisture and light conditions, and is less affected by crop residual coverage, so it can adapt to different soil environment which realize the non-destructive monitoring of sunflower seedlings.