这是一个基于yolov8s改进的无人机检测项目，具体改进事项和细节后续会逐渐补齐
This is a yolov8s based on the improvement of the drone inspection project, the specific improvements and details will be gradually fill in the subsequent

We use mmyolo as the development framework, you can refer to the official website of mmyolo at https://github.com/open-mmlab/mmyolo
-------------------------------------------------------------------------------------------------

configuration file：configs/custom_dataset
Outcome document：work_dirs/

-------------------------------------------------------------------------------------------------
这里给出一些简单的常用代码示例：
Some simple examples of commonly used code are given here:
train：
python tools/train.py "configuration file path"
such as:
python tools/train.py configs/custom_dataset/yolov8_s_200e_drone.py

test:
python tools/test.py "configuration file path" "Training weight paths"
suchu as
python tools/test.py configs/custom_dataset/yolov6_s_syncbn_fast_8xb8_40e_voc.py work_dirs/yolov6_s_syncbn_fast_8xb8_40e_voc/epoch_30.pth 

-------------------------------------------------------------------------------------------------
predict:
python demo/image_demo.py "image path" "configuration file path" "Training weight paths" "Save Path for Prediction Results"
such as:
python demo/image_demo.py demo/2007_007415.jpg ./configs/custom_dataset/yolov6_s_syncbn_fast_8xb8_40e_voc.py ./work_dirs/yolov6_s_syncbn_fast_8xb8_40e_voc/epoch_30.pth --out-dir ./demo/output

-------------------------------------------------------------------------------------------------
FPS测试：
FPS test:
python tools/analysis_tools/benchmark.py "configuration file path" "Training weight paths"
such as
python tools/analysis_tools/benchmark.py configs/custom_dataset/yolov8_s_syncbn_fast_4x8b_30e_voc.py work_dirs/yolov8_s_syncbn_fast_4x8b_30e_voc/best_coco_bbox_mAP_epoch_25.pth

-------------------------------------------------------------------------------------------------
打印模型的 Flops 和 Parameters，并以表格形式展示每层网络复杂度
Print the Flops and Parameters of the model and show the network complexity at each layer in tabular form
python tools/analysis_tools/get_flops.py "configuration file path"