Smartphone-Assisted Microscopic Imaging Prediction Of Mass And Chlorophyll Contents In Duckweeds
| dc.contributor.author | How, Mun Cheng | |
| dc.date.accessioned | 2022-10-12T23:27:19Z | |
| dc.date.available | 2022-10-12T23:27:19Z | |
| dc.date.issued | 2022-07-01 | |
| dc.description.abstract | The macroscopic image analysis technique can be utilized to quantify various physical features of plants with good accuracy, which has been proven by various researchers in recent years. Since microscopic imaging allows a more detailed observation of the plant structures compared to macroscopic imaging, investigation on the possibility of estimating the mass and chlorophyll content of three different species of duckweeds, namely Spirodela polyrhiza, Lemna minor and Wolffia arrhiza with smartphone captured microscopic images were carried out in this study. The most suitable magnification for correlation of mass with area for Spirodela polyrhiza, Lemna minor and Wolffia arrhiza were 1.2x, 1.2x and 1.5x respectively. The results showed good correlation between mass and area of plantlet of each species with R2 value of higher than 0.9 at the most suitable magnification. On the other hand, the best magnification for the correlation of chlorophyll content with color parameters and combinations was the same for all the three species which was 2.5x. Relatively low R2 value (< 0.7) was obtained for the correlation of chlorophyll content with each color parameter and combination with linear regression model. Non-linear regression model with Artificial Neural Network (ANN) modelling showed better fitting than linear regression model as significantly higher R2 value (> 0.8) was obtained for each model developed. The R+G+B model (R2 = 0.9029; RMSE = 1.1506 mg/L) was the best model to predict chlorophyll content in Spirodela polyrhiza whereas the B model was the best model for both Lemna minor (R2 = 0.9033; RMSE = 0.3375 mg/L) and Wolffia arrhiza (R2 = 0.9624; RMSE = 1.1180 mg/L). The mass of duckweed plantlets could be predicted well with linear regression model while the prediction of chlorophyll content in duckweed plantlets could be performed using developed ANN models with high accuracy. | en_US |
| dc.identifier.uri | http://hdl.handle.net/123456789/16303 | |
| dc.language.iso | en | en_US |
| dc.publisher | Universiti Sains Malaysia | en_US |
| dc.title | Smartphone-Assisted Microscopic Imaging Prediction Of Mass And Chlorophyll Contents In Duckweeds | en_US |
| dc.type | Other | en_US |
Files
License bundle
1 - 1 of 1
Loading...
- Name:
- license.txt
- Size:
- 1.71 KB
- Format:
- Item-specific license agreed upon to submission
- Description: