[bcd90f]: / MODEL / Validation.py

Download this file

146 lines (115 with data), 5.5 kB

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
import os
import cv2
from ultralytics import YOLO
import numpy as np
import pandas as pd
from sklearn.metrics import classification_report, accuracy_score, f1_score
from tqdm import tqdm
# Paths to your images and labels folders
images_folder = r'D:\CSE\Semester - 5\Extra\IIT jammu\YOLO\Code\Data\valid\images'
labels_folder = r'D:\CSE\Semester - 5\Extra\IIT jammu\YOLO\Code\Data\valid\labels'
# Initialize lists to store results
true_labels = []
predicted_labels = []
# Initialize lists to store IoU scores
iou_scores = []
# Initialize the YOLOv8 model
model = YOLO("best_75.pt")
# Initialize lists to store detection results
detection_results = []
def calculate_iou(box1_tensor, box2_list, image_width, image_height):
# Extract values from the tensor
x1, y1, x2, y2, confidence, class_id = box1_tensor.tolist()
# Convert values from the list to strings
class_id2, x2_gt, y2_gt, w2_gt, h2_gt = map(str, box2_list)
# Convert coordinates to float
x2_gt, y2_gt, w2_gt, h2_gt = float(x2_gt), float(y2_gt), float(w2_gt), float(h2_gt)
# Convert the model's bounding box coordinates to YOLO format
x1, y1, x2, y2 = x1 / image_width, y1 / image_height, x2 / image_width, y2 / image_height
# Calculate the intersection coordinates
xA = max(x1, x2_gt)
yA = max(y1, y2_gt)
xB = min(x1 + x2, x2_gt + w2_gt)
yB = min(y1 + y2, y2_gt + h2_gt)
# Calculate the intersection area
interArea = max(0, xB - xA) * max(0, yB - yA)
# Calculate the areas of the boxes
box1Area = x2 * y2
box2Area = w2_gt * h2_gt
# Calculate the IoU
iou = interArea / float(box1Area + box2Area - interArea)
return iou
# Iterate through image files and corresponding label files
for image_file in tqdm(os.listdir(images_folder)):
if image_file.endswith('.png'):
image_path = os.path.join(images_folder, image_file)
# Determine class (bleeding or non-bleeding) from image file name
is_bleeding = image_file.startswith('img-')
# Load the image
image = cv2.imread(image_path)
# Perform inference
results = model.predict(image)
# Check if bleeding is detected
detected_bleeding = results[0].boxes.shape[0] > 0
# Append true and predicted labels
true_labels.append(is_bleeding)
predicted_labels.append(detected_bleeding)
# Extract ground truth bounding boxes from label file
label_file_name = os.path.splitext(image_file)[0] + '.txt'
label_file_path = os.path.join(labels_folder, label_file_name)
if os.path.exists(label_file_path):
with open(label_file_path, 'r') as label_file:
lines = label_file.readlines()
ground_truth_boxes = [list(map(float, line.strip().split())) for line in lines]
# Calculate IoU for each detected box
if results[0].boxes.shape[0] > 0:
for box in results[0].boxes.boxes:
iou_values = []
for gt_box in ground_truth_boxes:
iou = calculate_iou(box, gt_box, 224, 224)
iou_values.append(iou)
max_iou = max(iou_values) if iou_values else 0
iou_scores.append(max_iou)
# Append detection results
if results[0].boxes.shape[0] > 0:
box = results[0].boxes[0]
detection_results.append({
'image_path': image_path,
'is_bleeding': is_bleeding,
'detection_confidence': box.conf[0].item()
})
# Calculate classification metrics
accuracy = accuracy_score(true_labels, predicted_labels)
recall = classification_report(true_labels, predicted_labels, target_names=['Non-Bleeding', 'Bleeding'], output_dict=True)
f1 = f1_score(true_labels, predicted_labels)
# Create a table of achieved evaluation metrics
classification_metrics = {
'Accuracy': [accuracy],
'Recall (Non-Bleeding)': [recall['Non-Bleeding']['recall']],
'Recall (Bleeding)': [recall['Bleeding']['recall']],
'F1-Score': [f1]
}
classification_df = pd.DataFrame(classification_metrics)
print("Classification Metrics:")
print(classification_df)
print()
print("Detection Metrics:")
print("IoU Metrics:")
# Calculate and print IoU statistics
iou_scores = np.array(iou_scores)
iou_mean = np.mean(iou_scores)
iou_median = np.median(iou_scores)
iou_75th_percentile = np.percentile(iou_scores, 75)
print("IoU Mean:", iou_mean)
print("IoU Median:", iou_median)
print("IoU 75th Percentile:", iou_75th_percentile)
# Calculate Detection Average Precision (AP) and Mean Average Precision (mAP)
detection_results_df = pd.DataFrame(detection_results)
detection_results_df['detection_confidence'] = detection_results_df['detection_confidence'].apply(lambda x: x.item() if isinstance(x, np.float32) else x)
detection_results_df['is_detected'] = detection_results_df['detection_confidence'] > 0.5 # You can adjust the confidence threshold as needed
detection_results_df['is_detected'] = detection_results_df['is_detected'].astype(int)
detection_ap = detection_results_df.groupby('is_bleeding')['is_detected'].apply(lambda x: np.mean(x))
mAP = detection_ap.mean()
print("Detection Average Precision (AP):")
print(detection_ap)
print("Mean Average Precision (mAP):", mAP)