Identification of Guava Fruit Shape (Psidium guajava L.) with Android-Based Digital Image Processing

Fruit shape determine the quality of guava which in turn affects greatly consumers’ appeal. Guava fruit is classified into intact and non-intact and is sorted manually using human eye which is less efficient and takes time. Digital image processing can be develop to effectively determine shape index of guava fruits. The objective of this study is to investigate the physical shape of red guava fruit using digital image processing, with a focus on roundness, sphericity, and K-value as shape indices. The calculation of these three parameters is done manually and image analysis. The results showed that manual roundness values ranged from 0.814 to 0.953. Image roundness values range from 0.871 to 0.908. Manual sphericity values range from 0.712 to 0.995. Image sphericity values range from 0.748 to 0.840. Manual K values range from 1.697 to 2.278. The image K value ranges from 1.631 to 3.285. Based on the results of the research, it is concluded that guava fruit can only be distinguished by the sphericity and K value shape index, due to the limited shape of the sample used so that it is not very visible the difference between intact and non-intact fruit with the roundness index.