Failed to Inference with a custom model

[Joe Lan]

Hi,
when I try to inference with a custom model using this code, error occur.

my code as follow:

`
import pixellib
from pixellib.instance import custom_segmentation

segment_image = custom_segmentation()
segment_image.inferConfig(num_classes=2, class_names=["BG", "butterfly", "squirrel"])
segment_image.load_model("./model/Nature_model_resnet101.h5")
segment_image.segmentImage("sample1.jpg", show_bboxes=True, output_image_name="sample_out.jpg")
`

error info:
Traceback (most recent call last):
File "D:/DownProjects/Pixellib/mytrain.py", line 38, in
test1()
File "D:/DownProjects/Pixellib/mytrain.py", line 35, in test1
segment_image.segmentImage("sample1.jpg", show_bboxes=True, output_image_name="sample_out.jpg")
File "D:\ProgramFiles\anaconda\envs\pixellib\lib\site-packages\pixellib\instance.py", line 565, in segmentImage
new_img = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
cv2.error: OpenCV(4.5.2) C:\Users\runneradmin\AppData\Local\Temp\pip-req-build-t9hleyt8\opencv\modules\imgproc\src\color.cpp:182: error: (-215:Assertion failed) !_src.empty() in function 'cv::cvtColor'

Thank you!

[ayoolaolafenwa]

@JoeLan96 the image you specify in the function does not exist. The error is an opencv error for an empty directory. Confirm the directory of your image.

[Joe Lan]

Thank you for your reply

I'm very sorry I made such a low-level mistake.

Now, I am going to train a sample model that does not need good results to segment the QR code in the picture. I want to know how many pictures need to be labeled as a training set.

Thanks in advance