Objective This study examines the evolution of social group opinions in physical space to understand the critical role of spatial factors in opinion dynamics. Methods By placing individual opinions on a square grid and distinguishing the bounded confidence Hegselmann -Krause HK model?? s trust matrix into local and non -local trust matrices a spatially bounded confidence HK model is constructed. By selecting different initial opinions and trust matrices with varying influence weights the evolution of group opinions in both information - isolated societies and networked information societies can be simulated. Results Numerical simulations reveal that when only the local trust matrix is active and initial opinions are sequentially arranged individual opinions first form local consensus before achieving global consensus. This simulates the evolution of group opinions in an information - isolated environment.Conversely when initial opinions are randomly arranged individual opinions do not form local consensus but instead evolve directly into global consensus. This simulates the evolution of group opinions in an information society. Conclusion The study demonstrates that under the combined influence of local and non-local trust matrices with the same influence weight a larger trust matrix value accelerates the reduction of the maximum number of opinion clusters. The influence weight reflects the scope of information dissemination its magnitude can either expedite or delay the formation of group opinion evolution as well as control the number of opinion clusters.