In the field of non-destructive testing of underground structures, the transient electromagnetic method has been widely used and highly recognized in applications such as grounding network corrosion defect detection due to its outstanding characteristics, including non-excavation operation and minimal terrain constraints. However, existing transient electromagnetic detection systems still face some urgent issues and limitations in terms of detection accuracy and adaptability to different environments. In order to overcome the above shortcomings, this paper innovatively investigated a transient electromagnetic grounding network defect detection method based on improved Newton iteration, and integrates this method with scanning imaging technology to accurately assess the structural information status of the grounding network. First of all, a theoretical model is established for transient electromagnetic detection through theoretical derivation, and quantitatively analyze the relationship between detection signal sensitivity and grounding network. Subsequently, the resistivity inversion is converted into a least-squares solution problem. The forward problem of the inversion process is calculated using an analytical theoretical model, while the inverse problem is implemented through improved Newton iteration method to estimate the equivalent resistivity. Finally, finite element numerical simulation was used to analyze the effects of soil, excitation signal cutoff time, and grounding grid corrosion on detection signals and inversion methods. Meanwhile, a 3×3 grounding network experimental platform was constructed to reconstruct three typical defects, demonstrating the accuracy and feasibility of the proposed method.