Design of Artificial Intelligence for Evaluation of In Vitro Potato (Solanum tuberosum) Microtuber Growth from Tissue Culture Based on Digital Imagery

Abstract The goal of this research is to introduce an AI model for assessing the development and yield of in vitro potato microtubers using tissue culture technology. The AI system was created to assist lab personnel in making decisions regarding microtubers produced through tissue culture. Previous research has indicated that the growth of microtubers is significantly impacted by the intensity of light. Once microtubers have formed, the light intensity needs to be decreased until it becomes dark, to facilitate optimal microtuber growth. The AI system takes in a digital image of the plant and other relevant information provided by the laboratory assistant. The digital image is then automatically processed to extract specific color intensity data for the microtubers and leaves. This intensity data is used as input for the fuzzy interference engine, which combines it with existing rules in the database to generate new rules. Additionally, the AI system employs a forward chaining mechanism to provide recommendations for appropriate actions that laboratory assistants should take based on the condition of a particular microtuber. Trial results demonstrate that the system’s recommendations are quite satisfactory, with an average accuracy of 95% in handling recommendations for potato microtubers.