To be effective, the assessment of earthquake disaster loss must be timely and reliable with respect to the real-time situation of the earthquake disaster area, improve the reaction capacity of the victims of destructive earthquakes, and meet the requirements of the emergency command. To improve the accuracy of assessments of earthquake disaster loss, remote-sensing images are processed by filtering and local gray-level adjustment. In this paper, we propose the identification of earthquake damage loss based on the template matching of remote-sensing images. To do so, remote-sensing images of the earthquake zone are acquired, then template matching is conducted with respect to the deviations between the acquired and template images. The eigenvalue of the feature region is extracted, and the imaging values of the spatial grid points are obtained. In this way, the earthquake disaster loss can be automatically evaluated according to the referenced imaging points, whereby the points with maximum energy indicate where the earthquake damage loss is greatest. Using this approach, the damage characteristics of the whole earthquake disaster area can be determined and extracted. The simulation experiment results show that the proposed method can extract the earthquake disaster loss and provide a preliminary reference for its assessment, although further study is needed to improve the general applicability of the proposed method.