摘要：【中文】本发明公开了一种图像处理和虚拟空间构建方法、装置、系统和存储介质。该方法包括：确定第一图像中的特征点和第二图像中的特征点，第一图像和第二图像是通过图像采集设备在不同拍摄点采集的物体对象的全景图像；对第一图像和第二图像进行特征匹配，确定第一图像和第二图像之间符合特征匹配条件的特征点，构成多个特征点对；利用多个特征点对，确定图像采集设备在每个拍摄点的运动姿态，根据运动姿态，确定不同拍摄点的相对位置；基于运动姿态、不同拍摄点的相对位置和多个特征点对，构建物体对象的三维模型。根据本发明实施例提供的图像处理方法，可以精确构建三维立体环境。 【EN】The invention discloses an image processing and virtual space construction method, device, system and storage medium. The method comprises the following steps: determining characteristic points in a first image and characteristic points in a second image, wherein the first image and the second image are panoramic images of an object acquired at different shooting points through image acquisition equipment; performing feature matching on the first image and the second image, determining feature points which accord with feature matching conditions between the first image and the second image, and forming a plurality of feature point pairs; determining the motion posture of the image acquisition equipment at each shooting point by utilizing the plurality of characteristic point pairs, and determining the relative positions of different shooting points according to the motion postures; and constructing a three-dimensional model of the object based on the motion attitude, the relative positions of the different shooting points and the plurality of characteristic point pairs. According to the image processing method provided by the embodiment of the invention, the three-dimensional environment can be accurately constructed.