As a transition zone between the city and the countryside, the urban fringe area is not only a geographical space affected by both of the regions, but also an area shrouded in conflicts of interest. The rapid development of urbanization witnesses tremendous changes the urban spatial structure is undergoing. Therefore, studying the spatial scope of the urban fringe area is conducive to the assessment of the current situation of urban development, and is further significant for policy formulation, population management, and resource allocation in the urban fringe area. Thanks to the development of remote sensing and geographic information technology, the types, quality, and accuracy of geospatial data that are applied to depict the characteristics of the urban fringe area have been significantly enhanced. Considering this, this paper takes the spatial mutation characteristics of the urban fringe area as a starting point, and a method for mutation point groups detection, combining multi-indices fusion and wavelet transform, is adopted to distinguish the spatial extent of urban fringe area based on the optimal results by grid-scale evaluation. And then we use the Delaunay triangulation automatic edge length threshold to extract the boundary of the mutation point groups and to obtain the spatial range of the urban fringe area. Empirical analysis is conducted taking Jiangyin City as the research area. The main experimental steps are as follow: firstly, several basic data are selected, containing land use data, road data, night light data, and service-oriented POI data according to the analysis of the characteristics of the urban fringe area. These multi-source data are then standardized based on grids and entropy weighting method for weight determination. In this way, the eigenvalues of the discriminant index of the urban fringe area are calculated. Secondly, based on the spatial mutation characteristics, the wavelet transform is employed to extract the mutation point groups on the eigenvalue sequence, which can effectively improve the discrimination accuracy of the mutation point group. It is suitable for the non-closed circular urban fringe area and for avoiding the influence of human subjective factors. Then, the algorithm based on the Delaunay triangulation automatic edge length threshold is utilized to extract the boundary of the mutation point groups and to obtain the spatial range of the urban fringe area, which can provide a reference for optimizing the urban spatial layout. Finally, the extraction results in this paper are compared with those obtained by the classical methods, like the information entropy model and the comprehensive index model of land use degree. It is apparent to see that the results from classical methods are more chaotic and scattered, while the results in this paper are more complete and objective. Comparing the extraction results of this paper with the extraction results of different index methods employing land use data and administrative division statistical yearbook data to construct urban fringe identification indicators for mutation detection, this study discovers the overlap between them is 88.03%, which testifies the factualness of this method. In terms of the analysis of local details, the results of this paper are more in line with the actual situation. To verify the effectiveness of the method proposed in this paper, the landscape pattern indices are adopted to test the range of urban built-up area, urban fringe area and rural hinterland extracted by the method of this paper and the other different index method respectively. Considering the patch class size landscape pattern indices, the areas delineated by two methods are following the spatial characteristics. Meanwhile, the value of data calculated in the urban fringe area identified in this paper is all higher than the counterpart method given the landscape size landscape pattern indices, the patch density, maximum patch index, landscape separation degree, landscape fragmentation index, and Shannon diversity index. The spread and Shannon uniformity, however, are both lower than the comparison method. It can be indicated that the fragmentation and heterogeneity of the landscape in the urban fringe area identified in this paper is higher, the landscape distribution is uneven, and the socio-economic conditions are more complex, thus proving the effectiveness of this method is especially suitable for the extraction of non-closed circular urban fringe area.