Estimation of Banana Fruit Maturity using Image Processing in MATLAB

Abstract

In the agriculture industry, maintaining a good quality product is important as the product will be marketed in the industry. Nowadays, fruit planting such as banana fruit has been produced in Malaysia’s economy. The process of banana fruit sorting and degrading based on its maturity was made before the fruit was marketed into the industry. This work focuses on the estimation maturity of banana fruit using an image processing technique where it can detect the ripeness of banana fruit. The technique used in this work is based on color image segmentation, image enhancement, detection of RGB values and removal of background images. The work considers a set of banana fruit images as samples that are processed using image processing in MATLAB software. A Graphical User Interface (GUI) was developed as interfacing between users and MATLAB. From the analysis of this study, the technique of image processing is able to detect different sample images of banana fruit and estimate the ripeness. This system is able to separate different conditions of banana fruit based on its maturity by detecting the skin color of the banana. The average RGB values of banana fruit are calculated and compared to identify the intensity of RGB on the sample images. The range of average RBG values of the banana fruits shows the intensity of the average value for red (R) is more than 250 pixels, for green (G) is 190-230 pixels and for yellow (Y) is lower than 120 pixels. In this study, a total of 25 sample image that includes 10 yellow, 10 green, 5 yellow and green banana fruit are used where a total of 21 sample images was detected. The accuracy of the system performance is calculated by using the total sample image detected. The accuracy of yellow banana sample images detected is 80%, for green banana is 90% and for the mix of green and yellow banana sample images is 80%. The overall result calculation of system accuracy detection banana fruit sample images is 84% successful rate. Development of this system able to help farmers to estimate the maturity of banana fruit without any damage.