Abstract
A user-friendly platform is developed to help potential users (i.e. researchers, students, and industry players who do not have or have little programming knowledges and modelling skills) to analysis spectral data so that they can predict the soluble solids content (SSC) of pineapple based on a highly over-lapping and colinearity visible and shortwave near infrared spectral data. First, users need to upload spectral data and their respective SSC data of pineapples to the platform via GUI functions. Second, two pre-processing techniques are available for users to eliminate the effects from unwanted signals. Third, users can start training their predictive model after choosing their desired number of inputs and complexity of the predictive model. After complete the training process, predicted SSC can be generated in seconds based on new spectral data.
A user-friendly platform is developed to help potential users (i.e. researchers, students, and industry players who do not have or have little programming knowledges and modelling skills) to analysis spectral data so that they can predict the soluble solids content (SSC) of pineapple based on a highly over-lapping and colinearity visible and shortwave near infrared spectral data. First, users need to upload spectral data and their respective SSC data of pineapples to the platform via GUI functions. Second, two pre-processing techniques are available for users to eliminate the effects from unwanted signals. Third, users can start training their predictive model after choosing their desired number of inputs and complexity of the predictive model. After complete the training process, predicted SSC can be generated in seconds based on new spectral data.
Current Problem/Background
Recently, Malaysia targets to move up beyond its 11th place ranking of the world¿s pineapple exporter by increasing its export value beyond the RM54mil mark this year via improved technologies and production. Due to the fact that a high nutrition and a good tasting are the crucial factors for the benefits of commercial, the pineapple internal quality assessment will be an area that both technology and market section concern about in this country. Among the internal quality attributes of pineapples, soluble solids content (SSC) is probably the most important factor to affect consumers¿ interest in pineapple purchasing. This is because there is a strong relationship between the eating quality of pineapple and SSC values. However, most of the available techniques for SSC measurements are time consuming, destructive, and difficult to be operated without a professional training and knowledge. A non-contact visible and shortwave near infrared spectroscopic techonology is widely used to acquire representative data that can be used to predict the SSC of pineapple. However, the information of these data are overlapping and highly correlated to each other. As a consequence, it is impossible to interprete them via direct inspection.
Recently, Malaysia targets to move up beyond its 11th place ranking of the world¿s pineapple exporter by increasing its export value beyond the RM54mil mark this year via improved technologies and production. Due to the fact that a high nutrition and a good tasting are the crucial factors for the benefits of commercial, the pineapple internal quality assessment will be an area that both technology and market section concern about in this country. Among the internal quality attributes of pineapples, soluble solids content (SSC) is probably the most important factor to affect consumers¿ interest in pineapple purchasing. This is because there is a strong relationship between the eating quality of pineapple and SSC values. However, most of the available techniques for SSC measurements are time consuming, destructive, and difficult to be operated without a professional training and knowledge. A non-contact visible and shortwave near infrared spectroscopic techonology is widely used to acquire representative data that can be used to predict the SSC of pineapple. However, the information of these data are overlapping and highly correlated to each other. As a consequence, it is impossible to interprete them via direct inspection.
Source: rmc.utm.my/inatex
0 comments:
Post a Comment