Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Highlights

Identification of large seed plantule point direction were realized using machine vision.

Abstract. A consistent plantule point and a uniform distribution of cucurbit seeds are required for factory seedlings and grafting. However, precision sowing Cucurbita ficifolia seeds automatically with a positioning and orientation method is difficult due to their large, non-spherical shape. To overcome this challenge, a machine vision-based precision seeder was designed specifically for Cucurbita ficifolia seeds. To achieve consistent seed placement and orientation at a 45° angle, a method using machine vision and precise seed positioning and orientation has been proposed. The plantule point location was determined by comparing the pixel values at both ends of seed. The positioning and orientation module was developed to achieve seeding with same orientation of plantule point and located in the center of the plug tray. Finally, the positioning and orientation precision seeder for seeds were tested and evaluated. The results showed that the accuracy rate of seed plantule point identification and the success rate of the seeds orientation actuator were 97.56% and 100.00%. 95.21% of seeds achieved a deflection angle of less than 15°. The germination port of 96.06% of the seeds is distributed in a circle with the center of the hole as the center and a radius of 10 mm. The qualified rate of positioning and orientation reached 87.80%, and the sowing efficiency was about 3750 grains/h. The results showed that the seeder achieves the consistent orientation of the germination port of seeds and sows the seeds in the center of the hole. This finding provides technical support for seedling testing and large-scale factory seedlings.