I. Introduction
Natural disasters are characterized by their suddenness, often causing a significant number of casualties within a short period, such as the 2023 Turkey earthquake [1]. Diagnosing and prioritizing treatment for casualties within a brief timeframe represents a systematic approach to field injury classification. However, the rapid classification and rescue of casualties at Casualty Collection Points (CCPs) often pose significant challenges due to the limited number of rescue personnel available. Search and rescue (SAR) robots hold promise in replacing rescue personnel for diagnosing injuries and rescuing casualties [2]. For instance, there is current research on SAR robots applied in underground building environments [3] and military casualty rescue [4].