Failed to load yolov3-spp.cfg and yolov2.cfg

System information (version)

OpenCV =>4.4
Operating System / Platform => Windows 64 Bit
Compiler => Visual Studio 2017

Detailed description

I tried some models from https://github.com/AlexeyAB/darknet，as listed bellow:	模型	模型大小
YOLOv2	194 MB	-
YOLOv2-tiny	43MB	15.3
YOLOv3	236 MB	1.3
YOLOv3-tiny	33.7 MB	13.3
YOLOv3-tiny-prn	18.8 MB	17.1
YOLOv3-SPP	240 MB	-
csresnext50-panet-spp-original-optimal_final	217 MB	1.8
YOLOv4	245 MB	1.2
YOLOV4-tiny	23.1 MB	9.8
enet-coco	18.3 MB	4.9

It failed to load yolov3-spp and yolov2 using the same code as other models uesd.

Steps to reproduce

The code is listed bellow:

import cv2
import matplotlib.pyplot as plt
import time
import numpy as np
coco_names = r"F:\opencv\sources\samples\dnn\coco.names"

model_yolov2 = r"F:\opencv\sources\samples\dnn\yolov2.weights"
cfg_yolov2 = r"F:\opencv\sources\samples\dnn\yolov2.cfg"

model_yolov2_tiny = r"E:\MachineLearning\darknet\yolov2-tiny.weights"
cfg_yolov2_tiny = "E:\MachineLearning\darknet\cfg\yolov2-tiny.cfg"

model_yolov3 = r"F:\opencv\sources\samples\dnn\yolov3.weights"
cfg_yolov3 = r"F:\opencv\sources\samples\dnn\yolov3.cfg"

model_yolov3_tiny = r"E:\MachineLearning\darknet\yolov3-tiny.weights"
cfg_yolov3_tiny = "E:\MachineLearning\darknet\cfg\yolov3-tiny.cfg"

model_yolov3_tiny_prn = r"E:\MachineLearning\darknet\yolov3-tiny-prn.weights"
cfg_yolov3_tiny_prn = "E:\MachineLearning\darknet\cfg\yolov3-tiny-prn.cfg"

model_yolov3_spp = r"F:\opencv\sources\samples\dnn\yolov3-spp.weights"
cfg_yolov3_spp = r"F:\opencv\sources\samples\dnn\yolov3-spp.cfg"

csresnext_model = r"E:\MachineLearning\darknet\csresnext50-panet-spp-original-optimal_final.weights"
csresnext_cfg = r"E:\MachineLearning\darknet\cfg\csresnext50-panet-spp-original-optimal.cfg"

model_yolov4 = r"F:\opencv\sources\samples\dnn\yolov4.weights"
cfg_yolov4 = r"F:\opencv\sources\samples\dnn\yolov4.cfg"

model_yolov4_tiny = r"E:\MachineLearning\darknet\yolov4-tiny.weights"
cfg_yolov4_tiny = "E:\MachineLearning\darknet\cfg\yolov4-tiny.cfg"

enet_model = r"E:\MachineLearning\darknet\enetb0-coco_final.weights"
enet_cfg = r"E:\MachineLearning\darknet\cfg\enet-coco.cfg"

img_file = r"C:\Users\admin\Pictures\car.jpg"
video_file = r'F:/opencv-4.4.0/samples/data/vtest.avi'

model = model_yolov3_spp
cfg = cfg_yolov3_spp

with open(coco_names,'rt') as f:
    names = f.read().rstrip('\n').split('\n')

def det_image_v1(model, cfg, img_file, c_threshold=0.5, nms=0.5):

    classes = names
    # initialize a list of colors to represent each possible class label
    COLORS = np.random.randint(0, 255, size=(len(classes), 3),dtype="uint8")

    print("[INFO] loading model...")
    net = cv2.dnn.readNetFromDarknet(cfg, model)

    # load the input image and construct an input blob for the image
    # by resizing to a fixed 300x300 pixels and then normalizing it
    image = cv2.imread(img_file)
    (H,W) = image.shape[:2]

    # Get the names of output layers
    ln = net.getLayerNames()
    ln = [ln[i[0] - 1] for i in net.getUnconnectedOutLayers()]

    # generate blob for image input to the network
    blob = cv2.dnn.blobFromImage(image,1/255,(416,416),swapRB=True, crop=False)
    net.setInput(blob)

    start = time.time()

    layersOutputs = net.forward(ln)
    print(layersOutputs)

    boxes = []
    confidences = []
    classIDs = []

    for output in layersOutputs:
        # loop over each of the detections
        for detection in output:
            # extract the class ID and confidence (i.e., probability) of
            # the current object detection
            scores = detection[5:]
            classID = np.argmax(scores)
            confidence = scores[classID]

            # filter out weak predictions by ensuring the detected
            # probability is greater than the minimum probability
            if confidence > 0.5:

                box = detection[0:4]* np.array([W, H, W, H])
                (centerX, centerY, width, height) = box.astype("int")

                # use the center (x, y)-coordinates to derive the top and
                # and left corner of the bounding box
                x = int(centerX - (width / 2))
                y = int(centerY - (height / 2))

                # update our list of bounding box coordinates, confidences,
                # and class IDs
                boxes.append([x, y, int(width), int(height)])
                confidences.append(float(confidence))
                classIDs.append(classID)

    # Remove unnecessary boxes using non maximum suppression
    idxs = cv2.dnn.NMSBoxes(boxes, confidences, c_threshold, nms)

    if len(idxs) > 0:
        # loop over the indexes we are keeping
        for i in idxs.flatten():
            # extract the bounding box coordinates
            (x, y) = (boxes[i][0], boxes[i][1])
            (w, h) = (boxes[i][2], boxes[i][3])

            # draw a bounding box rectangle and label on the image
            color = [int(c) for c in COLORS[classIDs[i]]]
            cv2.rectangle(image, (x, y), (x + w, y + h), color, 2)
            text = "{}: {:.4f}".format(classes[classIDs[i]], confidences[i])
            cv2.putText(image, text, (x, y - 5), cv2.FONT_HERSHEY_SIMPLEX,
                0.4, color, 1)

    end = time.time()
    # print the time required
    print('FPS:', 1/(end- start))

    # show the output image
    cv2.imshow("Image", image)
    cv2.waitKey(0)
    cv2.destroyAllWindows()

det_image_v1(model,cfg, img_file)

def det_image_v2(model, cfg, img_file, c_threshold=0.5, nms=0.5):
    # 加载yolo模型
    net = cv2.dnn_DetectionModel(model, cfg)
    net.setInputSize(512,512) # 设置网络输入尺寸
    net.setInputScale(1.0/255)
    net.setInputSwapRB(True)

    frame=cv2.imread(img_file)

    classes, confs, boxes = net.detect(frame, c_threshold, nms)

    for id, conf, box in zip(classes.flatten(), confs.flatten(), boxes):
        label = '{}, {:.2f}'.format(names[id], conf)
        # print(label)
        labelsize, baseLine= cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX,0.5,1)
        left, top, width, height = box
        top = max(top, labelsize[1])
        cv2.rectangle(frame, box, color=(0, 255, 0), thickness=3)
        cv2.rectangle(frame, (left, top-labelsize[1]),
                     (left+labelsize[0], top+baseLine),(255, 255, 255), cv2.FILLED)
        cv2.putText (frame, label,(left, top), cv2.FONT_HERSHEY_SIMPLEX, 0.5,(0,0, 0))

    # plt.imshow(frame[:,:,::-1])
    # plt.show()
    cv2.imshow('frame', frame)
    cv2.waitKey(0)
    cv2.destroyAllWindows()

det_image_v2(model, cfg, img_file)

I used cv2.dnn.readNetFromDarknet() and cv2.dnn_DetectionModel()， both of them failed. The error is:

error: OpenCV(4.4.0) C:\Users\appveyor\AppData\Local\Temp\1\pip-req-build-2b5g8ysb\opencv\modules\dnn\src\darknet\darknet_importer.cpp:207: error: (-212:Parsing error) Failed to parse NetParameter file: F:\opencv\sources\samples\dnn\yolov3-spp.cfg in function 'cv::dnn::dnn4_v20200609::readNetFromDarknet'

Issue submission checklist

[ ] I report the issue, it's not a question
[ ] I checked the problem with documentation, FAQ, open issues, answers.opencv.org, Stack Overflow, etc and have not found solution
[ ] I updated to latest OpenCV version and the issue is still there
[ ] There is reproducer code and related data files: videos, images, onnx, etc

opencv / opencv

Failed to load yolov3-spp.cfg and yolov2.cfg #18216

System information (version)

Detailed description

Steps to reproduce

Issue submission checklist