Potato machinery has become more intelligent thanks to advancements in autonomous navigation technology. The effect of crop row segmentation directly affects the subsequent extraction work, which is an important part of navigation line detection. However, the shape differences of crops in different growth periods often lead to poor image segmentation. In addition, noise such as field weeds and light also affect it, and these problems are difficult to address using traditional threshold segmentation methods. To this end, this paper proposes an end-to-end potato crop row detection method. The first step is to replace the original U-Net’s backbone feature extraction structure with VGG16 to segment the potato crop rows. Secondly, a fitting method of feature midpoint adaptation is proposed, which can realize the adaptive adjustment of the vision navigation line position according to the growth shape of a potato. The results show that the method used in this paper has strong robustness and can accurately detect navigation lines in different potato growth periods. Furthermore, compared with the original U-Net model, the crop row segmentation accuracy is improved by 3%, and the average deviation of the fitted navigation lines is 2.16°, which is superior to the traditional visual guidance method.

Aimed at solving the problem of the incidence of large film-breaking holes in the current sweet potato boat-bottom-shaped transplant, a boat-bottom-shaped up-film transplanting device was designed. Starting from the agronomic requirements of sweet potato up-film transplanting with a boat-bottom posture, an arc-shaped gripping mechanism was designed. In order to avoid further tearing and breaking the film hole during the movement of the clamping mechanism under the membrane, a transposition mechanism and a positioning cam were designed to jointly control the movement of the clamping mechanism. In order to ensure the stability of the transplanting plant spacing during the operation of the machine, a rotary encoder was used to collect the forward speed of the machine and adjust the speed of the power shaft in real time. Field experiments were carried out in order to verify the reliability and stability of the operation of the above-mentioned design mechanism. In the field test, the advance speed was used as the test factor, and the test indexes included three indexes of the seedlings after transplanting and three indexes of the film holes. The test results showed that the qualified rate of node depth in the middle of the seedling (QRNDMS) is between 95.67% and 86.79%, the qualified rate of root depth at the base of the seedlings (QRRDS) is between 94.00% and 90.11%, and the qualified rate of seedling length into the soil (QRSLS) is between 97.89% and 91.67%. The film hole spacing is between 240.33 mm and 245.32 mm, the film hole length is between 37.01 mm and 42.10 mm, and the film hole length variation rate is between 1.05% and 7.48%. This study can serve as a reference for up-film transplantation, particularly for boat-bottom-shaped up-film transplantation devices.

A transplanting manipulator and a transplanting conveyance device were devised to address the problems with current sweet potato transplanters, which cannot properly finish boat-bottom-shaped transplanting and have a high seedling damage rate. The ground wheel transmits power to the two devices in a specific transmission ratio to ensure their synchronous movement and avoid accumulated deviations. We selected the transplanting manipulator’s parameter combination that satisfies the optimal transplanting trajectory by the VB program. To avoid damage to the seedlings during clamping, the clamping force is only provided by the spring through the structural design. Combined with the agronomic requirements of boat-bottom-shaped transplanting and the characteristics of soil resistance during the transplanting process, the transplanting depth and crank rotation speed were selected as test factors. The qualified rates of node depth in the seedling middle (QRNDSM), the qualified rates of seedling root depth (QRSRD), and the qualified rates of stalk length into the soil (QRSLS) were established, and Design-Expert was used for the quadratic regression and orthogonal combination tests to optimize the factors. The results of the verification tests reveal that when the transplanting depth is in the range of 45.56 to 58.26 mm and the crank rotation speed is in the range of 39.39 to 43.45 rpm, the maximum QRNDSM is 94.91%, the maximum QRSRD is 95.42%, and the maximum QRSLS is 94.93%, which satisfies the industry standards. Research into sweet potato transplanting can benefit from this work, especially in the direction of sweet potato boat-bottom-shaped transplanting.