红外图像盲元自适应检测及补偿算法

Algorithm of blind pixels adaptive detection and compensation for infrared image

基于场景的盲元检测与补偿算法能有效地克服固定和随机盲元造成的图像亮点与暗点,提高探测器成像质量。通过将待检测图像分解为盲元目标和一般背景,建立迭代修正的背景预测模型提高预测准确度,并根据盲元和正常像元的残差能量统计差异完成单帧图像盲元检测。依据盲元缓变特征,利用建立的背景预测模型和盲元检测结果设计补偿算法,实现了单帧和序列图像的快速盲元补偿。结果表明:该算法对孤立和连续盲元均能准确检测,校正效果好,是一种有效、实用和易于软、硬件实现的在线检测及补偿算法。

Scene-based blind pixels detection and compensation algorithm can effectively remove the white or black （or grey） points in the image caused by fixed and random blind pixels, and improve the imaging quality. The infrared image was decomposed into blind pixels targets and normal pixels background. An iterative correction algorithm of background prediction model was established to improve predicting accuracy, single frame blind pixels were detected based on which the residual error difference among blind and normal pixels. Then, according to blind pixels gradual feature, the background prediction model and blind pixels detection result, a blind pixels compensation algorithm was proposed, which was applied to correct the infrared image and sequences. The experiments demonstrate that the algorithm has the advantages of good versatility for blind pixels location, good correction effect and easy implementation for software and hardware.