Search

Fruit cells are living irregular three‐dimensional (3D) transparent objects which makes them challenging to determine their real 3D size and shape through only two‐dimensional (2D) images using the existing biological microscope. This study deals with a newly self‐developed biological microscope including a microscope imaging system, a light source system, a stage and a support base for the 3D size characterization of fruit single cells. The main design concept is based on two optical path systems set up at the front (x‐axis) and bottom (z‐axis) directions of a transparent chamber containing single cells that allow the front view and bottom view of the single cell to be observed. Performance indicators such as mass, size, observation range, objective magnification, total magnification, focal range, focal accuracy, and resolution of the developed biological microscope were estimated. Finally, the 3D geometry size of single tomato cells was measured by the new biological microscope to demonstrate the relative ease at which accurate real 3D geometry information of single fruit cells could be obtained, which echoes its scientific value. [ABSTRACT FROM AUTHOR]

Non-line-of-sight (NLOS) target detection utilizing multipath plays an important role in anti-terrorism, urban warfare, indoor rescue and intelligent driving. In this paper, an imaging method based on single-channel ultra-wideband (UWB) radar and dual-view observation is proposed for detecting NLOS targets. First, the cause of producing ghosts from a single view in a previous study is analyzed in detail. On this basis, with focusing on dual-view observation, the ghost problem is handled at two levels. From the level of the radar echo, path partition is conducted with the utilization of dual-view observation, solving the problem of uneven energy among multipaths. After that, two images are generated by multipath imaging, respectively, which reduces the accumulation of ghosts. From the level of the image, owing to the distinct distribution of ghosts between the imaging result, the ghosts at different locations are eliminated by image fusion. Experimental results demonstrate that compared with single view and direct multipath imaging, the approach based on dual-view observation successfully eliminates most ghosts while retaining the targets in both single-target and double-targets scenarios, which verifies the effectiveness of the method. [ABSTRACT FROM AUTHOR]

Non-line-of-sight (NLOS) target detection utilizing multipath plays an important role in anti-terrorism, urban warfare, indoor rescue and intelligent driving. In this paper, an imaging method based on single-channel ultra-wideband (UWB) radar and dual-view observation is proposed for detecting NLOS targets. First, the cause of producing ghosts from a single view in a previous study is analyzed in detail. On this basis, with focusing on dual-view observation, the ghost problem is handled at two levels. From the level of the radar echo, path partition is conducted with the utilization of dual-view observation, solving the problem of uneven energy among multipaths. After that, two images are generated by multipath imaging, respectively, which reduces the accumulation of ghosts. From the level of the image, owing to the distinct distribution of ghosts between the imaging result, the ghosts at different locations are eliminated by image fusion. Experimental results demonstrate that compared with single view and direct multipath imaging, the approach based on dual-view observation successfully eliminates most ghosts while retaining the targets in both single-target and double-targets scenarios, which verifies the effectiveness of the method.