Method for implementing image processing of wedge-and-strip photon counting detector by FPGA

Abstract

The invention belongs to the technical field of photoelectric detection and relates to a method for implementing image processing of a wedge-and-strip photon counting detector by the FPGA. The method includes the steps: a CPU writes coordinate calculation formula parameters into an image processing unit; the image processing unit receives image data from an image collection system; the coordinate calculation formula parameters are transformed, a floating-point number is enlarged into an integer expressed in a hexadecimal form; formula simplification is performed by transformation; a division method for a coordinate calculation formula is subjected to reduction of fractions to a common denominator, and division amount is reduced; multiplication of parameters of the coordinate calculation formula and image data is parallelly implemented by an IP core; results of the multiplication are subjected to summation or subtraction according to symbols of the parameters, and a sum or a difference is used as a dividend for division; division of the coordinate calculation formula is achieved through the IP core, and a quotient and a remainder are stored respectively; the calculation remainder and coordinate calculation constants are multiplied, and final coordinates are obtained by rounding and dividing the coordinate calculation constants and are transmitted to the CPU for image storage.

Description

technical field [0001] The invention relates to a method for realizing image processing of a photon counting detector, belonging to the technical field of photoelectric detection. Background technique [0002] The photon counting detector is a detector that converts the incident photon collection into multiple charge signals. The charge signal is shaped and amplified, peak held and sampled by the signal processing circuit to obtain image data. The image processing is based on the coordinate calculation formula to convert the image data Convert to photon coordinate value, and then perform cumulative counting at the position corresponding to the coordinate value to generate the final photon counting image. The highest incident photon frequency that can be responded to by the photon counting detector, signal processing circuit, and image processing three-part series is called the photon count rate, and the minimum displacement of the photon coordinates that can be resolved is c...

AI Technical Summary

Problems solved by technology

[0003] There are two commonly used floating-point calculation methods: the first is to use ground operating system software, such as VC, Labview, this method is simple and flexible, and can meet the requirements of image processing speed and resolution, but cannot Translated into engineering applications, only suitable for ground tests

The second is to use embedded software, such as DSP. This method is conducive to engineering transformation and is suitable for aerospace requirements. However, coordinate calculations are usually floating-point operations, and there are many multiplication and division operations, while embedded software is serial processing. way, resulting in a longer image processing operation cycle, the operation speed can be improved by increasing the embedded software system clock, but it not only increases the design difficulty but also reduces the system reliability, and the ability to improve the operation speed is limited, when the photon counting rate is high , still cannot meet the requirement

Images