Watermelon maturity level based on rind colour as categorization features

Abstract

This paper proposed categorization of watermelon maturity level based on rind colour. 30 samples are used to develop the new algorithm. Firstly, image of the fruit is captured in RGB format using a digital camera. The RGB image then is converted into gray scale image to undergo pre-processing. Here, the watermelon is extracted from the unwanted background by scanning the image for connected pixels or blobs, followed by the elimination of the unwanted blobs having certain area, leaving the desired object intact. Masking process is performed to obtain image of the watermelon with a black background. The colour components of the masked image namely red, green and blue are extracted and used as features in determining the maturity level of the watermelon. Another 30 samples are used to test the technique to validate the proposed method. Initial results attained proved that this method is appropriate and at par as compared to opinion from the human expert.

Determination of the Ripeness State of Guavas Using an Artificial Neural Network

“…The classification is performed through an integration and analysis of the color variation in red-green. Using variation of color, in [14] an algorithm is developed to determine the ripeness state of watermelons. Another method used in classification tasks is the fuzzy logic, for example in [6] color features and texture features are extracted from images of tomatoes and they are submitted to an if-then rules system to get a classification.…”