After the training process has been completed, when an image has been given as input for recognition there are bounding boxes all over the image, whereas it must be only on the ball since the model is trained only to recognize the ball.
Please help resolve this issue.
After the training process has been completed, when an image has been given as input for recognition there are bounding boxes all over the image, whereas it must be only on the ball since the model is trained only to recognize the ball. Please help resolve this issue.
![img2](https://user-images.githubusercontent.com/31180335/61686777-858afa00-acd5-11e9-9f71-e28f299eac2f.PNG)