高灵敏面阵红外相机实测图像盲元检测与补偿方法

Blind Element Detection and Compensation Method for High Sensitivity Area Array Infrared Camera Tested Images

摘要: 使用面阵红外相机，可大大提高成像视场角，使大视场高空间分辨率红外成像成为可能，但是高辐射分辨率面阵相机具有高灵敏度，在进行挂飞实验时，会实时的产生盲元，凭借硬件升级无法完全避免盲元产生的问题，因此对红外图像有效进行盲元检测和补偿成为需要研究的课题。本文提出一种改进的基于场景的盲元检测方法，通过对连续时域红外图像求均匀背景图像得到差值图像，结合三梯度检测法，利用空域灰度中值和梯度绝对值之间的关系，设置阈值对盲元进行判别，在进行盲元补偿后得到处理后的图像。实验结果证明该方法效果明显，能有效地检测到连续的盲元块，检测更加准确，在后续进行盲元补偿时，图像质量也进一步提升。

Abstract: The use of a planar array infrared camera can greatly improve the imaging field Angle and make it possible to create infrared imaging with a large field of view and high spatial resolution. However, a planar array camera with high radiation resolution has high sensitivity, and blind elements will be generated in real time during the flight test. Therefore, effective blind element detection and compensation of infrared images have become a research topic. In this paper, an improved scene-based blind element detection method is proposed. The difference image is obtained by obtaining uniform background image of continuous time-domain infrared image. Combined with the three-gradient detection method, the relationship between the spatial median gray value and the absolute value of the gradient is used to set a threshold for blind element discrimination, and the processed image is obtained after blind element compensation. The experimental results show that the method is effective and can effectively detect continuous blind blocks, and the detection is more accurate. The image quality is further improved when the blind element compensation is carried out later.