Green hydrogen plays a pivotal role in decarbonizing our energy system and achieving the Net-Zero Emissions goal by 2050. Offshore wind farms (OWFs) dedicated to green hydrogen production are currently recognized as the most feasible solution for scaling up the production of cost-effective electrolytic hydrogen. However, the cost associated with distribution and transmission systems constitute a significant portion of the total cost in the large-scale wind-hydrogen system. This study pioneers the simultaneous optimization of the inter-array cable routing of OWFs and the location and capacity of offshore hydrogen production platforms (OHPPs), aiming to minimize the total cost of distribution and transmission systems. Considering the characteristics of hydrogen production efficiency, this paper constructs a novel mathematical model for OHPPs across diverse wind scenarios. Subsequently, we formulate the joint planning problem as a relaxed mixed-integer second-order cone programming (MISOCP) model and employ the Benders decomposition algorithm for the solution, introducing three valid inequalities to expedite convergence. Through validation on real-world large-scale OWFs, we demonstrate the validity and rapid convergence of our approach. Moreover, we identify hydrogen production efficiency as a major bottleneck cost factor for the joint planning problem, it decreases by 1.01% of total cost for every 1% increase in hydrogen production efficiency.