Collision accidents pose a serious threat to the navigation and operational safety of fishing vessels. To enhance the decision-making capability of fishing vessel operators, this paper proposes an intelligent collision avoidance decision-making method based on the Convention on the International Regulations for Preventing Collisions at Sea (COLREGS). This method determines the encounter situation of the vessel in accordance with the COLREGS requirements and calculates the degree of collision risk, represented by the Collision Risk Index (I_CR), based on the Distance to the Closest Point of Approach (d_CPA) and the Time to the Closest Point of Approach (t_CPA). By investigating the collision avoidance behaviors of multiple fishing vessel operators, the Goodwin marine observation results were revised to establish the Safe Distance Approach (S_DA) specific to fishing vessels. Furthermore, the traditional Artificial Potential Field (APF) method was improved and integrated with a Genetic Algorithm (GA) to ensure that fishing vessels can avoid collisions safely and effectively. Finally, simulation experiments were conducted to validate the effectiveness of the proposed algorithm. The results demonstrate that the proposed method overcomes the limitations of the single artificial potential field approach in previous vessel collision avoidance decision-making systems. It achieves optimal steering amplitude and determines the appropriate timing for returning to the original course, thereby providing a safe and collision-free decision-making solution that complies with COLREGS. This study can serve as a valuable reference for practical collision avoidance operations by deck officers.