Prosperity in the military and national defense is an essential indicator of a nation's comprehensive strength. The division's member nations support both peace and development, but this does not imply that their military capabilities are underdeveloped. In recent years, the military has employed deep learning, an advanced artificial intelligence technique, for a variety of purposes, including automated driving, situational awareness, and data fusion on the battlefield. Situational awareness relies on object detection, and deep learning-based object detection techniques need a lot of data to function well. However, gathering a lot of data in combat situations is challenging and sensitive, making it challenging to train high-accuracy detectors that can be applied in real-world scenarios at the moment. To address this, we provide an MODD-λ dataset in a virtual environment for Land-Air Integration and Cross-Domain Collaborative Unmanned Swarm Systems object detection. To our best knowledge, we are the first to acquire datasets to serve land-air integration and cross-domain collaborative unmanned swarm systems in a Battlefield Game. Our research provides innovative perspectives on how to improve the efficiency of unmanned combat systems and reduce the cost of data acquisition, while supporting the state in implementing smarter military strategies.