超分辨率图像重建能突破硬件限制提高图像分辨率，是图像处理领域的研究热点。重建过程可以用含积分的平差模型来描述，因此可以用测量平差的方法来解决。由于这种含积分的非线性平差模型中观测值与待求函数之间是积分关系，待求的是函数而不是参数，目前的平差理论还不能处理这一模型。本项目将深入研究测量平差理论在超分辨率图像重建中的应用，研究超分辨率图像重建时相应平差模型的建立、解算及从平差的角度研究超分辨率图像重建时低分辨率图像之间的几何关系优化。具体研究如下：1.图像的灰度函数及超分辨率图像重建时相应的积分型非线性平差模型的建立；2.积分型非线性平差模型的具体解法及在超分辨率图像重建中的应用；3.像素作为观测值时的随机模型建立及平差权的确定；4.基于平差理论的低分辨率图像之间相对分布对重建效果的影响研究。有关研究能丰富测量平差的理论，并为超分辨率重建开辟新途径。

Super-resolution image reconstruction(SRIR) can break through hardware limitation to improve image resolution, which now is a research focus in the field of image processing. Because the reconstruction process can be described by adjustment model with integration, surveying adjustment can be used to solve the problems in SRIR. Owing the relationship between observed value and unknown function is integral relationship in the integral nonlinear adjustment model, and the unknown function must be found rather than parameters, presented adjustment theory can not deal with this model. This project will make further study on surveying adjustment theory in SRIR, study the establishment and calculation of adjustment model, and optimize the geometric relationship of low resolution images in reconstruction. The main contents are as follows: 1. The establishment of gray function and integral nonlinear adjustment model. 2. Calculation of integral nonlinear adjustment model and its application in SRIR. 3. Random model of pixel when it is taken as observion and determination of its weight. 4. The effect of geometric relationship between low resolution images on SRIR. This research will enrich surveying adjustment theory and provide a new method for SRIR.