Automated rubber seed clones identification using imaging technique and statistical analysis / Muhammad Fahmi Harun

This paper describe research work to recognize selected rubber seed series clones using image processing techniques based on perimeter, area and radius. There are five types of rubber seed clone from the species Hevea brasiliensis of rubber seed have been used as samples in this project which are RRIM2002, RRIM2015, RRIM2020, RRIM2023 and RRIM2024. Sample of rubber tree seeds are captured using digital camera where the RGB color image are stored and processed. Processing involves segmentation algorithm which includes thresholding and application of morphological technique to extract the shape features. Another 225 samples are used for testing and the analysis is done using SPSS software to identify the clones correctly. From the observed one-way ANOVA and error plot measurement, it shown that all of clones series significantly different from each other for perimeter classification but only two series shows significantly different for area and radius classification. As a conclusion, perimeter of rubber seed clone can be used in order to recognize all selected best rubber tree clones compare with area and radius that can only be used for RRIM2002 and RRIM2015.