Intelligent High-Performance Programmable Infrared Preprocessing SoC and ISP Algorithm Implementation

With the continuous advancement of infrared detection imaging technology and the increasing demand for applications, there is a need for integrated applications of infrared imaging (IR imaging) systems. The design of these applications aims to achieve programmability, modularity, and intelligence in IR imaging verification systems through a high-performance integrated hardware architecture. Therefore, in the present study, an infrared image (IR image)-processing system-level SoC was designed for integrated application in IR imaging systems. The chip integrates a CPU, an IR image-processing coprocessor, Cameralink image output, and dedicated peripheral control interfaces, with a maximum real-time data-processing capability of over 1.9 Gbps. The chip can realize real-time image processing such as infrared detector control, IR image acquisition, adaptive non-uniformity correction, adaptive blind pixel recognition and replacement, image equalization (enhancement), image stretching and shrinking, pseudocolor conversion, image flipping and mirroring, and image windowing, simplifying the structure of infrared detection and signal processing in the IR imaging system and laying a technical foundation for the implementation of the integrated design of IR imaging systems. The chip is implemented using 40 nm CMOS process technology, with a size of 4000 × 6800 μm, an operating temperature of 218.15~398.15 K, and a frame rate of 200 Hz. The test results show that, after processing with the chip, the non-uniformity and blind pixel rate of the image can be reduced to 0.13% and 0.15%, respectively; the image stretching and shrinking rate can reach 25%; the image can be windowed at any size; and the expected requirements are achieved in IR image processing. At the same time, the system’s power consumption is less than 0.5 W, its packaging weight is less than 0.4 g, and its volume is reduced by 75% compared with traditional systems, meeting the application requirements of systems with high volume, power consumption, cost, and performance requirements. Furthermore, the proposed system has high engineering application value and prospects, promoting the further development of infrared imaging technology and more intelligent IR image algorithms.