A method for characterizing the panicle traits in rice based on 3D micro-focus X-ray computed tomography

Characterizing panicle traits is essential for measuring rice production. Because these traits can be used to analyze the effects of different genes on the rice. The spikelet number and the seed setting rate are usually considered to be more crucial. In existing characterizing methods, each grain of the panicle must be spread out manually to reduce the overlapping areas and made existing 2D imaging for the unfolded panicle. However, these methods, which maybe influence the panicle structure, are inefficient and complicated for operating, and cannot determine the seed setting rate. For the problems mentioned above, X-ray CT, as a non-destructive technique, can be used to examine and characterize the internal structure of the panicle. Therefore, in this paper, the panicle was scanned and reconstructed by Microfocus CT System. The panicle traits were detected by combination of distance transform watershed algorithm and 3D connected domains. The effectiveness of this method was verified by accuracy calculation of eight different rice panicle samples. This method is of great significance in practical application.