Search

The discrete element method (DEM) for modeling the behavior of particulate material is highly dependent on the use of appropriate and accurate parameters. In this study, a seed metering DEM simulation was used to measure, calibrate, and verify the physical and interactional parameters of rapeseed. The coefficients of restitution and static friction between rapeseeds and three common materials (aluminum alloy, acrylic, and high-density polyethylene) were measured using free drop and sliding ramp tests, respectively. The angle of repose was determined using a hollow cylinder experiment, which was duplicated using a DEM simulation, to examine the effects of static and rolling friction coefficients on the angle of repose. Response surface optimization was performed to determine the optimized model parameters using a Box–Behnken design test. A metering device was made with three materials, and rapeseed seeding was simulated at different working speeds to verify the calibrated parameters. The validation results showed that the relative errors between the seed metering model and experiments for the single qualified seeding, missed seeding, and multiple seeding rates were −0.15%, 3.29%, and 5.37%, respectively. The results suggest that the determined physical and interactional parameters of rapeseed can be used as references for future DEM simulations.

The cultivation of hybrid rice requires lower investment cost and fewer transplanted seedlings per hill to improve the grain yield by increasing the productive tillers. Therefore, for the precise transplanting of hybrid rice, it is crucial to accurately recognize machine-transplanted seedlings and detect seedlings per hill. In this study, an automatic approach based on machine vision technology was proposed for detecting seedlings per hill. For the problem of rice field background interference on machine-transplanted seedling extraction, rice seedling and paddy field pixels were obtained from rice field images to generate two patches. The distribution characteristics of colour component values of these two patches under different colour models (i.e., RGB, YCrCb, HSV, Lab, HLS and LUV) were analysed by adopting exploratory factor analysis. These analysis results showed that the Lab (L-factor of brightness, a-content of red or green, and b-content of yellow or blue) colour model outperformed other models in separating seedlings from the background. The preferred Lab colour model along with Otsu’s method were used to extract rice seedling information, and the skeleton of the seedling hill was extracted using the thinning algorithm to effectively characterize the morphological structure of single seedling hill. An algorithm for detecting endpoints of skeletonized seedling hills was proposed to represent leaf tips of seedlings as endpoints, and the relationship between ground truth counts and automatic counts of endpoints was positive with R2 and root mean square error (RMSE) of 0.9105 and 0.7437, respectively. Combining the number of endpoints with the number of skeletons, a mathematical model was developed and used for detecting seedlings per hill of machine-transplanted hybrid rice. The overall detection accuracy of seedlings per hill of machine-transplanted hybrid rice was up to 93.5%. The processing time for detecting single seedling hill image was less than 50 ms. These results show that the proposed approach allowed for the effective, reliable and fast detection of seedlings per hill, which provides technical assistance for further precise adjustments of the machine-transplanted performance of hybrid rice.