CODEWITHSUNDEEP
pythonmodelobject-detectionyolov8
Sofia
Sofia

Reputation: 1

Object detection on python, what does the command "save_txt=True" return in the following code: "result= model('V3.mp4',save=True, save_txt=True)"?

In Google-Colaboratory (Python) I trained a model to detect the ball object in a video. I want to see the results of the model and I use the following command: result= model('V3.mp4',save=True, save_txt=True) In addition to saving me the video with the ball detection, I included the command "save_txt=True" because I also need to generate a .txt file with the coordinates of the individual frames detected by the model.

Within each frame, where only one ball was detected, there are the following number values: 0 0.777123 0.55625 0.0212264 0.0625

As a base model and weights to train my model I used yolov8.

I thank you for your help.

Can you tell me what exactly these five values indicate?

Upvotes: 0

Views: 156

Answers (1)

hanna_liavoshka
hanna_liavoshka

Reputation: 1323

The results here are returned in the YOLO Object Detection Dataset format:

class_id  x_center  y_center  width      height
0         0.777123  0.55625   0.0212264  0.0625

Box coordinates are in a normalized xywh format (from 0 to 1). To get these values in pixels, multiply x_center and width by image width, and y_center and height by image height.

More information about the detection results can be obtained from the documentation: https://docs.ultralytics.com/modes/predict/#working-with-results

Upvotes: 0

Related Questions