openpose_coco

OpenPose trained on COCO Pose dataset

Single pose human estimation.

Requirements

Run pip install -r requirements.txt

This will install opencv and numpy, it's only dependencies.

How to run

You'll need to use the file detector.py.

import cv2
import numpy as np
import sys
import random

from detector import BODY_PARTS, POSE_PAIRS, PoseDetector

PROTO = "/path/to/downloaded/prototxt/file"
MODEL = "/path/to/downloaded/caffemodel/file"

detector = PoseDetector(PROTO, MODEL)

img_path = "/path/to/image"
img = cv2.imread(img_path)

# The run method receives a numpy array with values [0-255] representing the image in [H, W, 3] shape
# `points` is a list of (x,y) representing all body parts.
points = detector.run(img)


for partFrom, partTo in POSE_PAIRS:
    idFrom = BODY_PARTS[partFrom]
    idTo = BODY_PARTS[partTo]

    if points[idFrom] and points[idTo]:
        cv2.line(img, points[idFrom], points[idTo], (0, 255, 0), 3)
        cv2.ellipse(img, points[idFrom], (3, 3), 0, 0, 360, (0, 0, 255), cv2.FILLED)
        cv2.ellipse(img, points[idTo], (3, 3), 0, 0, 360, (0, 0, 255), cv2.FILLED)

cv2.imshow('image',img)
cv2.waitKey(0)
cv2.destroyAllWindows()

The above script will show the annotated image.

Customize

Deal with the returning points object to do whatever you want!

Model info

• Using OpenPose official models from here

Comment

We are using openCV's amazing dnn library that allows us to load Tensorflow, Caffe and Torch models. You can go here for more information.