Combination algorithms applied to source reconstruction for neutron coded images and restoration for incomplete coded images

The neutron emission of compressed capsules filled with fuels in inertial confinement fusion implosions can be measured by neutron imaging systems. Source reconstruction is an important method in coded-aperture imaging. In this paper, we use a combination algorithm to reconstruct the neutron source image. This method can improve the resolution and signal-noise ratio of the reconstructed image. In addition, the ray tracing method is used to obtain the point spread functions of the whole field of view (250 µm), and thus, the system response can be obtained. The edge gray interpolation method is used to restore the missing portion of incomplete coded images. The method can maintain a good performance when the missing-data angle is limited to less than 50°.

Design of the scintillator imaging lens for the neutron imaging system at the 100 kJ-level laser facility

Deuterium-tritium neutron yield has reached up to about 10¹³ at the 100 kJ-level laser facility, which makes measurement of neutron emission images possible with the neutron imaging system. There are two methods to collect neutron images from the scintillator array, optical fiber taper and the lens system. Here, we report a design of the lens system for the neutron imaging system at the 100 kJ-level laser facility. The lens system, which consists of a nine-element collecting lens, with a spatial resolution of 20 µm and a light-collection efficiency of 5.9% has been designed.

A novel approach to correct the coded aperture misalignment for fast neutron imaging

Aperture alignment is crucial for the diagnosis of neutron imaging because it has significant impact on the coding imaging and the understanding of the neutron source. In our previous studies on the neutron imaging system with coded aperture for large field of view, "residual watermark," certain extra information that overlies reconstructed image and has nothing to do with the source is discovered if the peak normalization is employed in genetic algorithms (GA) to reconstruct the source image. Some studies on basic properties of residual watermark indicate that the residual watermark can characterize coded aperture and can thus be used to determine the location of coded aperture relative to the system axis. In this paper, we have further analyzed the essential conditions for the existence of residual watermark and the requirements of the reconstruction algorithm for the emergence of residual watermark. A gamma coded imaging experiment has been performed to verify the existence of residual watermark. Based on the residual watermark, a correction method for the aperture misalignment has been studied. A multiple linear regression model of the position of coded aperture axis, the position of residual watermark center, and the gray barycenter of neutron source with twenty training samples has been set up. Using the regression model and verification samples, we have found the position of the coded aperture axis relative to the system axis with an accuracy of approximately 20 μm. Conclusively, a novel approach has been established to correct the coded aperture misalignment for fast neutron coded imaging.