为提高红外图像目标场景细节表现能力, 研究了热图像自适应细节增强算法AS&DRP。通过能自适应调节窗口大小的自适应滤波器滤波及γ压缩变换, 在压缩像素数位的同时有效地保留了目标和场景的细节信息。采用Xilinx公司的XC5VSX50T的FPGA作为核心处理器, 配制少量外围器件研制了通用数字图像处理电路, 并在其上实现自适应细节增强算法。通过对线性压缩和由自适应细节增强算法处理图像的主观视觉评价与客观评价参数EME的对比, 表明了算法在突出细节上的优势。算法硬件处理系统具有时间延迟小、消耗资源有限、功耗低等优点, 适合在各类高性能热成像系统上使用。

To improve the capabilities of the performances of the target scene details got from the infrared device, the Adaptive smoothing and dynamic range partitioning compression algorithm AS&DRP is proposed. Through the adaptive filter that can adaptively adjust window size and γ compression, this algorithm compresses the gray levels the same time preserves the details. Setting XC5VSX50T of the Xilinx Company as the core processor, a general digital processing circuit with a number of peripheral devices to achieve adaptive smoothing and dynamic range partitioning compression algorithm is developed. Comparing the images of linear compression with the images of adaptive smoothing and dynamic range partitioning compression, either subjective feelings or objective evaluation parameters proves the superiorities of adaptive smoothing and dynamic range partitioning compression algorithm on detail enhancement. Meanwhile, this algorithm has the advantages of small delay、limited consumption of resources and so on. It’s suitable for various types of high-performance thermal imaging systems.