Image Processing Techniques for Harumanis Disease Severity and Weighting Estimation for Automatic Grading System Application

Abstract

Harumanis Mango is known as the king of Mangoes. It is very nutritious and rich with carotenes. However, many of the farmers and agriculture experts reported that they have problems in grading and inspecting the Harumanis Mango. Sometimes, Mango production loses its quality due to diseases that are not even visible to the naked eyes. Traditionally, farmers and agriculture experts will estimate the severity of the disease using their experiences. While for weight estimation, manual inspection was done by using a weight scale. This traditional method has its own drawbacks as it can lead to some errors due to inconsistencies made by human inspection. Furthermore, they are less efficient and very time-consuming. Therefore, an automated procedure that able to classify the disease severities and weight estimations would be much appreciated. With the aid of image processing techniques, diseases can be classified according to its scale, and its weight can be estimated. A number of pixels of Harumanis Mango will be used for classification. The analysis will be done by using the statistical method of regression. It shows that the accuracy of weight estimation is 72.25%.