一种新的棉花根系图像阈值分割方法

为了方便实现棉花原位根系的检测,提出了一种自动全局阈值分割方法,用于棉花根系图像分割.采用无损的数码设备成像法对棉花原位根系图像进行采集,使用自动全局阈值分割方法分割图像.该方法首先将采集到的原位根系图像进行空间转换,使采集到的图像在HSV空间下进行分割;然后,通过全局阈值分割的方法选择阈值将图像进行二值化处理,采用闭运算方法对二值图像进行初步降噪;最后,通过形状特征的筛选过滤图像中所有噪声,并筛选出图像中细长的根部特征.本算法可以有效抑制噪声和土壤杂质的影响,能够对根系进行准确地分割,是快速检测作物根系图像的有效办法.

A new threshold segmentation method for cotton root images

In order to detect the cotton root in situ, an automatic global threshold segmentation method was proposed for cotton root image segmentation. Digital device imaging method was used to collect the cotton in-situ root system images, and the automatic global threshold method was used to segment the images. Firstly, the acquired root images are spatially transformed, and the acquired root images are segmented in HSV space. Then, the global threshold segmentation method is used to select the threshold value to binarize the image, and the closed operation method is used to denoise the binary image. Finally, all the noises in the image were filtered through the shape feature screening, and the slender root features in the image are selected. This algorithm can effectively suppress the influence of noise and soil impurities, and can accurately segment the root system. It is an effective method for fast detection of crop root images.