Landmine contamination remains a critical barrier to post-conflict recovery, necessitating detection methods that are safe, scalable, and cost-effective. This study investigates drone-based detection of relevant plants in a mixed vegetation as required for future plant-based biodetection of explosive remnants derived from landmines. Field experiments were conducted in Ukraine and Denmark using oilseed radish and winter rapeseed as indicator biosensors due to their suitability for genetic modification and resilience in diverse climates. High-resolution RGB and multispectral imagery were collected under varying conditions, including flight altitude, time of day, and land preparation. We evaluated plant visibility at different altitudes and trained models to detect the plants and developed strategies to streamline annotation. Practical deployment challenges, such as seeding and growth of indicator plants in non-cultivated and hazardous soils are also considered. Our findings demonstrate the potential of aerial biosensor monitoring and offer insights into the development of plant-based landmine detection systems suitable for real-world applications.