Python api not labeling object by name

🤝 Community Help
1

Hi - I’m trying to use the web api from python on a mac to do object detection. I can get it to open a webcam and I can get it to process the images and it draws a box around the objects but doesn’t label the objects. In particular, it’s a different behavior than the webcam in the “Visualize” interface.

Curious how to get it to annotate what the objects are and have the same functionality as in the “Visualize” page. Are there “settings” per model that I need to change? (if so I can’t find them). Or is there a way to call the API so that it adds the detected object name?

Here is my code:

# load config
import json
with open('config.json') as f:
    config = json.load(f)

    ROBOFLOW_API_KEY = config["ROBOFLOW_API_KEY"]
    ROBOFLOW_MODEL = config["ROBOFLOW_MODEL"]
    ROBOFLOW_SIZE = config["ROBOFLOW_SIZE"]

    FRAMERATE = config["FRAMERATE"]
    BUFFER = config["BUFFER"]

import cv2
import base64
import numpy as np
import requests
import time

# Construct the Roboflow Infer URL
upload_url = "".join([
    "https://detect.roboflow.com/",
    ROBOFLOW_MODEL,
    "?api_key=",
    ROBOFLOW_API_KEY,
    "&format=image",
    "&stroke=5"
])

# Get webcam interface via opencv-python
video = cv2.VideoCapture(0)

# Infer via the Roboflow Infer API and return the result
def infer():
    # Get the current image from the webcam
    ret, img = video.read()

    # Resize (while maintaining the aspect ratio) to improve speed and save bandwidth
    height, width, channels = img.shape
    scale = ROBOFLOW_SIZE / max(height, width)
    img = cv2.resize(img, (round(scale * width), round(scale * height)))

    # Encode image to base64 string
    retval, buffer = cv2.imencode('.jpg', img)
    img_str = base64.b64encode(buffer)

    # Get prediction from Roboflow Infer API
    resp = requests.post(upload_url, data=img_str, headers={
        "Content-Type": "application/x-www-form-urlencoded"
    }, stream=True).raw

    # Parse result image
    image = np.asarray(bytearray(resp.read()), dtype="uint8")
    image = cv2.imdecode(image, cv2.IMREAD_COLOR)

    return image

# Main loop; infers sequentially until you press "q"
while 1:
    # On "q" keypress, exit
    if(cv2.waitKey(1) == ord('q')):
        break

    # Capture start time to calculate fps
    start = time.time()

    # Synchronously get a prediction from the Roboflow Infer API
    image = infer()
    # And display the inference results
    cv2.imshow('image', image)

    # Print frames per second
    print((1/(time.time()-start)), " fps")

# Release resources when finished
video.release()
cv2.destroyAllWindows()
2

For your specific question of showing the class name on the detections, you can set the labels query parameter to true which would display the labels on the predictions. See our Hosted API docs for Object Detection

Alternatively, since you’re using Python, you can also use Supervision as well

3

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.