Hi,
why does retina sometimes revert my pictures 90 degres when detecting object? It doesnt detect almost anything when reverted...
Thx, JS
Post the full code you ran here...
from imageai.Detection import ObjectDetection
import os
from IPython.display import Image
from PIL import Image
import numpy as np
os.chdir(...)
input_path = ...
tmp = os.listdir(input_path)
for j in tmp:
input_img = str(j)
execution_path = os.getcwd()
detector = ObjectDetection()
detector.setModelTypeAsRetinaNet()
detector.setModelPath(os.path.join(execution_path, "resnet50_coco_best_v2.0.1.h5"))
detector.loadModel()
custom_objects = detector.CustomObjects(person=False, car=False, bottle=True)
main_obj, detections, extracted_obj = detector.detectCustomObjectsFromImage(
input_image=os.path.join(execution_path, input_path+input_img),
input_type="file",
# output_type="file",
output_type='array',
output_image_path=os.path.join(execution_path, "im"),
custom_objects=custom_objects,
extract_detected_objects=True,
minimum_percentage_probability=25)
for eachObject in detections:
print(eachObject["name"], " : ", eachObject["percentage_probability"], " : ")
print("--------------------------------")
os.mkdir(input_img)
img = Image.fromarray(main_obj, 'RGB')
img.save(input_img + '/main.png')
#img.show()
cntr = 0
for i in extracted_obj:
try:
img = Image.fromarray(i, 'RGB')
img.save('/Users/jakubstech/Documents/detekce/' + input_img + '/extracted' + str(cntr) + '.jpg')
except ValueError:
print('error')
cntr += 1`
@OlafenwaMoses ? :)
Kindly install the latest version of ImageAI
pip3 install imageai --upgrade
Did it solve the issue? Any news on this?
I think this should be closed to keep things clean. Of course it can be reopened later
cc @OlafenwaMoses
Most helpful comment
I think this should be closed to keep things clean. Of course it can be reopened later
cc @OlafenwaMoses