研究目的
To correct spatial nonuniformity in infrared imaging systems for variable integration time and time-varying offset with a one-time calibration method.
研究成果
The proposed two-dimensional calibration method effectively corrects spatial nonuniformity for variable integration times and time-varying offsets in infrared imaging systems. It eliminates the need for multiple calibrations or shutters, providing real-time correction with low computational complexity. Experimental results show superior performance compared to traditional two-point correction, with reduced FPN and higher SNR across a range of integration times, making it suitable for practical engineering applications.
研究不足
The method assumes a linear response model and may not fully account for all nonlinearities in the infrared camera response. It relies on the accuracy of the calibration process and the stability of the FPA temperature, which is stabilized by a cooler but could still have minor drifts. The attenuation factor's effectiveness depends on the calibration conditions, and performance might degrade under extreme variations not covered in calibration.
1：Experimental Design and Method Selection:
A two-dimensional calibration scheme is designed to use images captured at different integration times and temperatures to compute correction coefficients. The method involves subtracting base images to cancel offset and applying gain and offset corrections based on the calibration data.
2：Sample Selection and Data Sources:
Calibration images are captured using a blackbody at temperatures of 20°C and 40°C, with integration times of 16 μs and 1.5 ms. Real scene images are captured for uniform targets and outdoor scenes with varying integration times.
3：5 ms. Real scene images are captured for uniform targets and outdoor scenes with varying integration times.
List of Experimental Equipment and Materials:
3. List of Experimental Equipment and Materials: A 256x256 cooled mid-wave infrared (MWIR) camera with a custom-designed readout circuit, column-parallel ADCs, integration time controller, and FPGA-based image processing circuits is used. A blackbody is employed for calibration.
4：Experimental Procedures and Operational Workflow:
Calibration involves capturing three images at specified temperatures and integration times, computing correction coefficients, and then applying these coefficients to real-time images by subtracting a base image (captured with short integration time) from the scene image and correcting the difference.
5：Data Analysis Methods:
Spatial nonuniformity is evaluated using fixed pattern noise (FPN) and spatial signal-to-noise ratio (SNR) metrics, calculated from the standard deviation and mean of pixel outputs.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
