The recorded field of view of any digital imaging system is limited by the physical size of the sensor array. While it is a common knowledge that the image field exists beyond the boundaries of the sensor array, there is generally no way to measure it unless the imaging system (or the object) is translated laterally. We propose a single-shot computational imaging system with a multiple-point impulse response that is able to effectively increase the physical size of the sensor array. The image recorded on the array sensor is not visually meaningful but can be used to recover the image field beyond the native sensor boundary via a sparsity based iterative algorithm, as we demonstrate in this work. The system concept can be considered analogous to structured illumination imaging; however, the structuring is performed here in the Fourier space in order to recover an extended image field. The effective increase in the sensor size depends on the extent of the impulses in the engineered multiple point impulse response. The achievable sensor size extension is therefore limited by the resolution of the phase mask that is introduced in the Fourier plane of the imaging system. We present a simulation study where the individual impulses in the designed impulse response extend over the original array sensor size, thereby doubling the effective sensor dimensions (a four times increase in the number of pixels) without affecting the image resolution. Both binary and gray-scale objects have been considered in our study in order to illustrate that the quality of the extended field of view image depends on the sparsity of the object under consideration. The concept of extended field of view computational imaging as presented here may find a number of practical applications.

任何数字成像系统的记录视野都受到传感器阵列物理尺寸的限制。众所周知，像场存在于传感器阵列的边界之外，除非成像系统（或物体）横向平移，否则通常无法对其进行测量。我们提出了一种具有多点脉冲响应的单次计算成像系统，能够有效地增加传感器阵列的物理尺寸。阵列传感器上记录的图像在视觉上没有意义，但可用于通过基于稀疏性的迭代算法恢复超出本地传感器边界的图像场，正如我们在这项工作中所展示的那样。可以认为系统概念类似于结构化照明成像；然而，此处在傅立叶空间中执行结构化以恢复扩展的图像场。传感器尺寸的有效增加取决于设计的多点脉冲响应中的脉冲程度。因此，可实现的传感器尺寸扩展受到在成像系统的傅立叶平面中引入的相位掩模的分辨率的限制。我们提出了一项模拟研究，其中设计的脉冲响应中的单个脉冲扩展到原始阵列传感器尺寸，从而在不影响图像分辨率的情况下将有效传感器尺寸加倍（像素数量增加四倍）。为了说明扩展视野图像的质量取决于所考虑对象的稀疏性，我们的研究中都考虑了二进制和灰度对象。这里提出的扩展视场计算成像的概念可以找到许多实际应用。