A Hardware Processing Method of Infrared Image Linear Interpolation Expansion

Abstract

The invention discloses an infrared image linear interpolation expanding hardware processing method. The infrared image linear interpolation expanding hardware processing method comprises the steps: establishing a coordinate mapping relation between pixel points of an expanded image and an original image; realizing algorithm design by adopting four-stage flow line of coordinate calculation, grayscale reading, expansion calculation and grayscale storage. A rapid addressing method without precision loss is adopted in the hardware design process, an FIFO controller realizes grayscale rapid reading method to guarantee efficient work of the flow line, the algorithm calculation speed is greatly increased, the hardware resource cost is reduced and the calculation precision is improved. The infrared image linear interpolation expanding hardware processing method is implemented under Xilinx virtex5-XC5VFX100T and operates at 100 M. The original image 160*128 is expanded by 1.2 times, namely 192*153, an image with the size being 1.1 times (namely 176*140) that of the original template image is captured from the expanded image, and algorithm calculation is finished only within 0.25 ms.

Description

technical field [0001] The invention relates to a hardware processing method for linear interpolation expansion of an infrared image, belonging to the field of special algorithm hardware circuit design. Background technique [0002] Information processing technology based on image matching has been widely used in the new generation of aerospace models, and the real-time performance of image processing directly affects the guidance accuracy. In the past, it is increasingly difficult to implement image processing algorithms based on general-purpose processors (such as DSP) by software to meet the real-time requirements of aerospace models. At present, the method of simplifying the algorithm and sacrificing the accuracy of the algorithm is often used to reduce the operation time of the algorithm. Algorithm acceleration through algorithm hardware is the most effective means to reduce algorithm operation time. [0003] At present, image processing adopts a matching method based...

AI Technical Summary

Problems solved by technology

[0003] At present, image processing adopts a matching method based on grayscale information, which has high matching accuracy but a relatively large amount of calculation.

Images