Abstract：To extract the center line of paddy seedling row, an algorithm was proposed based on segmented vertical projection. The binary color image after preprocessing was divided into three sections of lower， middle and upper in the longitudinal direction. From the lower section， the vertical projection method was used to determine the edge around the crop row.The feature points between the left and right edges were extracted to fit a temporary line, and the intersection position of the line and the top of the section was used to determine the edge points belonging to the same crop row in the middle section.The same method was used in the upper section of the image, and all of the feature points were extracted into different sets of points according to different crop rows. The least square method was used to fit each point in the point set representing the crop row. The experimental results show that in the 1 280×1 024 pixel scolor image,the accuracy of the proposed method is 9450%， and the average time of each frame is 208 ms.
