Impacts from falling objects can pose a major risk of failure for offshore pipelines. Appropriate protection measures are required to reduce this risk. To achieve this goal, a risk-based integrity model is proposed. Based on the surrogate model of genetic programming, the finite element analysis is cooperated with the probabilistic method to consider nonlinear effects on the failure probability. The required validations are performed to demonstrate the accuracy. The cumulative prospect theory is introduced to encapsulate the risk-attitudes in the consequence evaluation. This integrated model is further extended to determine an optimal protection strategy to obtain the maximum return on an investment, as well as ensure safety of the infrastructures. A case study is used to illustrate the methodology used to determine the burial depth, which is a primary protection measure. The parametric study showed that the soil property and pipeline wall thickness have a considerable influence on the risk. A comprehensive overview of the outputs from different risk-attitudes is also provided. This study can provide a reference for offshore pipeline safety design, and the proposed model can be used as a potential guideline for other researchers to develop customised risk assessment models.