共查询到20条相似文献,搜索用时 484 毫秒
1.
在红外成像制导应用中,为满足长周期免拆卸贮存的应用需求,红外导引头非均匀性的研究越来越多的集中于采用自适应的校正方法来代替传统的参考源的非均匀性校正方法。针对传统基于神经网络的自适应非均匀性校正算法容易造成"鬼影"的问题,提出了一种改进的红外导引头成像自适应非均匀性校正算法。该方法在传统神经网络非均匀性校正的基础上,进行了4点实用化的改进:首先,通过对图像运动判断,避免场景静止时的过学习;其次,采用自适应学习率,避免细节丰富区域的过学习;然后,利用双边滤波求期望目标的评估,减少细节的损失;最后,通过判断误差函数的波动量来决定是否对偏置进行更新。实验结果表明,该方法在校正精度、收敛速度和稳定性方面均优于传统的神经网络校正算法。 相似文献
2.
3.
4.
5.
6.
提出一种新的能对非均匀性校正效果进行定量计算的评估测度“校正率”,并进行了实验验证.校正率以时域噪声为参考标准来衡量图像非均匀特性,不仅可以反映图像显示效果,还可以反映图像测温精度.该方法可用于IRFPA系统的性能评价和非均匀性校正方法校正效果评价,对非均匀性校正的研究具有重要意义. 相似文献
7.
基于神经网络的红外焦平面非均匀校正的新算法 总被引:2,自引:1,他引:1
传统的神经元网络算法对噪声具有较好的自适应性,但当噪声略强时,它的校正效果会出现下降,为进一步提高性能,原作者提出了基于神经元网络的红外焦平面非均匀性校正的改进算法。但在场景静止时,原算法就不再适用。针对这种情况,分析了基于神经元网络的红外焦平面非均匀校正的改进算法,提出了在场景静止时的校正算法。并结合两者,最后提出了基于神经元网络的红外焦平面非均匀校正的新算法。仿真证明,新算法具有优异的性能。 相似文献
8.
9.
由于红外偏振焦平面的异构特性,在非均匀校正过程中需要考虑不同检偏通道的响应差异对整体校正效果的影响,其非均匀校正问题相较同构的普通红外焦平面更为复杂。针对红外偏振焦平面的非均匀校正问题,提出了一种基于场景偏振冗余估计的非均匀校正算法,通过对场景图像和由场景图像计算得到的偏振冗余估计图像进行统计,得到整个焦平面上所有像元响应在统计特性上的差异,然后分通道从两个方向对这些差异进行比较分析,得到更新后的增益校正系数,再通过辐射重定标抑制由于静止场景所造成的鬼影,得到当前状态下相机的增益校正系数。在这个过程中,通过偏振冗余估计评价之前的校正系数,自适应地实现增益校正系数的更新。最后使用真实场景数据进行测试,结果表明本文所提出的非均匀校正算法有效提高了所获取偏振图像的准确性。 相似文献
10.
基于场景的SPRITE热像仪的非均匀性数字校正 总被引:7,自引:2,他引:5
非均匀性校正是热成像信号预处理技术的重要内容。本文针对SPRITE热成像系统的成像特点,提出了一种基于场景统计特性的非均匀性数字校正算法,并给出了实现方法和仿真结果。 相似文献
11.
12.
针对当前采集到的场景出现缺损导致场景设计重建效率较低、准确性较差的问题,提出基于激光雷达扫描及关键点特征匹配的室内场景设计重建。通过Deleta 2B型激光雷达传感器直接采集室内场景图像,利用最小二乘拟合滤波算法曲线拟合室内场景图像的灰度值,根据图像特征向量匹配室内场景图像的关键点特征,通过德洛内三角化算法串联室内场景点云数据,构造三维网格模型,通过wallis滤波器提升重建辨识度,实现室内场景的重建。实验结果表明,该方法重建后的室内图像可辨识度较高,信噪比最高可达到498dB,室内场景图像覆盖率均在90以上,重建场景的结构相似度接近1,次卧室场景重建时间为592 h。 相似文献
13.
提出了一种结合图像匹配和神经网络算法的焦平面阵列非均匀性校正算法。算法首先用最新的校正系数对图像进行非均匀性校正,输出校正结果;然后对相邻两帧图像进行匹配,估计出相邻帧之间图像的运动量;最后用神经网络算法分别对校正系数进行正向和反向自适应更新。采用图像匹配技术保证了校正系数更新时不会引起场景的模糊,采用校正系数双向更新策略可以保证每帧都能对每个像元的系数至少进行一次更新,与常用的神经网络校正算法相比,降低了对场景统计特性的要求,收敛速度较快。使用模拟添加噪声和采集的红外图像序列对算法进行仿真验证,结果表明,给出的算法校正效果优于常用的神经网络非均匀性校正算法。 相似文献
14.
15.
The fusion of different resolution SAR images 总被引:20,自引:0,他引:20
Costantini M. Farina A. Zirilli F. 《Proceedings of the IEEE. Institute of Electrical and Electronics Engineers》1997,85(1):139-146
We explore the possibility of using different data sets relative to the same scene to obtain a better knowledge of the scene than the one obtained using only one data set. In particular we concentrate on the fusion of two different spatial resolution images, although the method we propose can be regarded as a method of more general interest. The fused image has the least mean square deviation from the finer resolution image, subject to the constraints imposed by the knowledge of the coarser resolution image. Fusion is obtained by solving a constrained quadratic highly parallelizable minimization problem. Explicit formulas for the solution of the minimization problem are given. The number of elementary operations required is proportional to the number of pixels of the finer resolution image. We test our method on a class of simulated images that reproduce some features of synthetic aperture radar (SAR) images. As an example we consider the problem of detection of point scatterers in a uniform background. The results obtained show that the information from the coarser resolution image can significantly improve the quality of the reconstructed scene obtained from the finer resolution image 相似文献
16.
17.
Underwater images often show severe quality degradation due to the light absorption and scattering effects in water medium. This paper introduces a scene depth regularized underwater image dehazing method to obtain high-quality underwater images. Unlike previous underwater image dehazing methods that usually calculate a transmission map or a scene depth map using priors, we construct an exponential relationship between transmission map and normalized scene depth map. An initial scene depth is first estimated by the difference between color channels. Then it is refined by total variation regularization to keep structures while smoothing excessive details. An alternating direction algorithm is given to solve the optimization problem. Extensive experiments demonstrate that the proposed method can effectively improve the visual quality of degraded underwater images, and yields high-quality results comparative to the state-of-the-art underwater image enhancement methods quantitatively and qualitatively. 相似文献
18.
Zhaoyi Jiang Zhangye Wang Qunsheng Peng 《Journal of Infrared, Millimeter and Terahertz Waves》2003,24(10):1737-1748
Although many models have been put forward to realize static infrared scene, they could not generate dynamic infrared scene real time in interactive way. In this paper a new method is proposed to solve the problem. We first model the targets and background of infrared scene based on the hybrid way of geometry and multi-spectral texture images. Then considering the attenuation effect of atmosphere and the noise mechanic of infrared image sensor, we present an infrared depth image model to generate dynamic images of the objects in the scene from different viewpoint. The complexity of infrared dynamic scene is thus reduced greatly and the reality of infrared scene is improved. Finally, real-time walkthrough for infrared scene is successfully realized and the average walkthrough speed is larger than 25 frames per second. 相似文献
19.
20.
高分辨率遥感图像地物信息丰富,但场景构成复杂,目前基于手工设计的特征提取方法不能满足复杂场景分类的需求,而非监督特征学习方法尽管能够挖掘局部图像块的本征结构,但单一种类及尺度的特征难以有效表达实际应用中复杂遥感场景特性,导致分类性能受限.针对此问题,本文提出了一种基于多尺度多特征的遥感场景分类方法.该算法首先设计了一种改进的谱聚类非监督特征(iUFL-SC)以有效表征图像块的本征结构,然后通过密集采样提取每幅遥感场景的iUFL-SC、LBP、SIFT等三种多尺度局部图像块特征,并通过视觉词袋模型(BoVW)获得场景的中层特征表达,以实现更为准确详实的特征描述,最后基于直方图交叉核的支持向量机(HIKSVM)进行分类.在UC Merced数据集以及WHU-RS19数据集上的实验结果表明本文方法可对遥感场景进行鉴别特征提取,有效提高分类性能. 相似文献