Rice stigma measurement using microscopic image analysis
2010
Recario, R.N.C., Mariano, V.Y., Philippines Univ. Los Baños, College, Laguna (Philippines). Inst. of Computer Science
Observing physical characteristics through counting, measuring and other means is an important element of research. Majority of plant research, particularly in rice research, employ the said activities. However, most of these activities are done manually. This leads to lag in productivity. The growing interest in the application of image analysis in rice research can aid the demands on further studies on rice. The study employs microscopic image analysis as a tool in analyzing rice stigmas through automated counting of pollens in the stigma and estimating the surface area of the rice stigma. The study is in collaboration with the International Rice Research Institute (IRRI) [Los Baños, Laguna, Philippines] who are studying if the surface area and number of pollens of rice stigma are significant indicators of stress (e.g. heat stress) applied on rice under study. To do that, digital images of rice stigma samples, taken using a microscope and magnified 50 times, were subjected to image processing methods such as thresholding, image cleaning, edge detection, object counting and shape from shading. Measurement of the rice stigma surface area would be in square micrometer (umsquare). Results of the automated method were compared with the results of manual counting and measuring. Over-all performances of the methods were evaluated in terms of accuracy of computed surface area and accuracy in counting pollens.
显示更多 [+] 显示较少 [-]