Study on the Image Processing Algorithm for Detecting the Seed-Sowing Performance

An image processing algorithm was developed for detecting a seed-sowing device's performance in this paper. Firstly, sequential images of the seeds on the conveyor belt of the test-bed were obtained using a camera by sensor triggering. These sequential images were then spliced into a single long image. Using discriminate analysis techniques, a threshold of the binary image was automatically obtained. According to the threshold, seeds could be abstracted from the background of the long image. The seed number on the binary image can be counted using the median as the single seed area (pixels number). Based on the information of longitudinal and transverse projection on the binary image, the breadth of the seed array, the coordinates of the seed array central and the distance between seed arrays, seed intervals and non-seed intervals of each seed array were obtained. Finally, for drill seeding, hill-drop seeding and precise seeding, more performance parameters could be further calculated based on the measurement parameters mentioned above.