请问yolo8是否考虑了图像长宽相差很大的情况

UNeedCryDear / yolov8-opencv-onnxruntime-cpp

yolov8 hub,cpp with onnxruntime and opencv

Apache License 2.0

319 stars 56 forks source link

Closed SuperGentry closed 1 year ago

SuperGentry commented 1 year ago

博主你好，比如有一张760x200的图片，把图片大小变为640x640输入模型推理的时候会不会出现预测内容偏移的情况？或者说，有没有必要先把图像大小提前填充为760x760的大小，然后再等比例放缩？

UNeedCryDear commented 1 year ago

代码里面会自动按照python源码的做法处理的，也就是先将长边resize到640，然后短边填充指定像素到640 https://github.com/UNeedCryDear/yolov8-opencv-onnxruntime-cpp/blob/4b90bf0c3d86d67faf1a3683ccace79163b7c16d/yolov8_utils.cpp#L14