Received: 06-01-2023 / Accepted: 27-01-2023
This study aimedto design an automaticmodelfor quality assessment of potatotuber seeds by applying computer vision. In the study, a number of sproutsandtuber scab diseases (Actinomyces scabiesand Helminthosporium solani) were selected to evaluate the quality of potatotuber seeds. Apicture of the object was taken by Pi 2/3 camera and sentit to Raspberry Pi 4B embedded computer to process the received image with the YOLO-v4 algorithm. Initially,the model drew conclusions about the quality of the seed potatoes through counting the number of sprouts and identifying scabdiseases on tubers with an average processing time of 0.147 seconds. The rate of accurate identification of the diseaseson infected tubers was 93.33% with Actinomyces scabiesand94.74% with Helminthosporium solani.95.56% of sprouted tubers were counted correctly. This evaluation model can be used in an automatic sorting potato seed tubers before planting.