datasetforge 0.1.0__tar.gz

datasetforge-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Shruti Shirpurkar
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

datasetforge-0.1.0/PKG-INFO

@@ -0,0 +1,14 @@
+Metadata-Version: 2.4
+Name: datasetforge
+Version: 0.1.0
+Summary: Computer Vision Dataset Creation Tool
+Author: Shruti Shirpurkar
+Requires-Python: >=3.8
+License-File: LICENSE
+Requires-Dist: opencv-python
+Requires-Dist: numpy
+Dynamic: author
+Dynamic: license-file
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary

datasetforge-0.1.0/README.md

@@ -0,0 +1,33 @@
+# DatasetForge
+
+DatasetForge is a Python library for creating, annotating, and managing image datasets from videos and webcams.  
+
+## Features
+- Convert videos into frame datasets
+- Annotate images with bounding boxes
+- Count objects in images
+- Preview datasets in grids
+- Capture images from webcam
+- Generate dataset statistics
+
+## Installation
+```bash
+pip install datasetforge
+
+##EXAMPLE USAGE
+import datasetforge as df
+
+# Convert video to dataset
+df.video_to_dataset("dataset/video.mp4", "dataset/video_frames")
+
+# View dataset stats
+df.dataset_stats("dataset/video_frames")
+
+# Annotate a single image
+df.annotate("dataset/video_frames/frame_0.jpg", save_folder="dataset/annotated")
+
+# Count objects
+df.count_objects("dataset/video_frames/frame_0.jpg")
+
+# Preview first 9 images
+df.preview("dataset/video_frames", max_images=9)

datasetforge-0.1.0/datasetforge/__init__.py

@@ -0,0 +1,3 @@
+from .video_dataset import video_to_dataset
+from .object_counter import dataset_stats
+from .annotator import annotate_frames

datasetforge-0.1.0/datasetforge/annotator.py

@@ -0,0 +1,49 @@
+import cv2
+import os
+import glob
+
+def annotate(image_path, save_folder="dataset/annotated"):
+    os.makedirs(save_folder, exist_ok=True)
+    save_path = os.path.join(save_folder, os.path.basename(image_path))
+
+    if os.path.exists(save_path):
+        img = cv2.imread(image_path)
+        cv2.putText(img, "Already annotated", (10, 30),
+                    cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 255), 2)
+        cv2.imshow("Annotation", img)
+        cv2.waitKey(500)
+        cv2.destroyAllWindows()
+        print(f"Skipping {image_path}, already annotated.")
+        return
+
+    img = cv2.imread(image_path)
+    if img is None:
+        print(f"Cannot read image: {image_path}")
+        return
+
+    while True:
+        r = cv2.selectROI("Select ROI (SPACE=OK, s=Skip, q=Quit)", img, False, False)
+
+        if cv2.waitKey(1) & 0xFF == ord('q'):
+            cv2.destroyAllWindows()
+            print("Quitting annotation.")
+            exit(0)
+        elif cv2.waitKey(1) & 0xFF == ord('s'):
+            cv2.destroyAllWindows()
+            print(f"Skipped {image_path}")
+            return
+        elif r[2] > 0 and r[3] > 0:
+            x, y, w, h = r
+            crop_img = img[y:y+h, x:x+w]
+            cv2.imwrite(save_path, crop_img)
+            cv2.destroyAllWindows()
+            print(f"Annotated saved: {save_path}")
+            return
+        else:
+            print(f"No ROI selected for {image_path}")
+            cv2.destroyAllWindows()
+            return
+
+def annotate_frames(frame_folder="dataset/video_frames", save_folder="dataset/annotated"):
+    for img_path in sorted(glob.glob(f"{frame_folder}/*.jpg")):
+        annotate(img_path, save_folder)

datasetforge-0.1.0/datasetforge/auto_annotate.py

@@ -0,0 +1,36 @@
+from ultralytics import YOLO
+import cv2
+import os
+
+model = YOLO("yolov8n.pt")
+
+def auto_annotate(image_path, label_folder="dataset/labels"):
+    os.makedirs(label_folder, exist_ok=True)
+
+    results = model(image_path)
+    img = cv2.imread(image_path)
+    h, w = img.shape[:2]
+
+    base = os.path.splitext(os.path.basename(image_path))[0]
+    label_file = os.path.join(label_folder, base + ".txt")
+
+    lines = []
+
+    for r in results:
+        boxes = r.boxes.xyxy
+        for box in boxes:
+            x1, y1, x2, y2 = box
+            bw = x2 - x1
+            bh = y2 - y1
+            xc = (x1 + bw/2)/w
+            yc = (y1 + bh/2)/h
+            bw /= w
+            bh /= h
+            lines.append(f"0 {xc} {yc} {bw} {bh}")
+
+    if lines:
+        with open(label_file, "w") as f:
+            for l in lines:
+                f.write(l+"\n")
+
+    print("Auto annotated:", image_path)

datasetforge-0.1.0/datasetforge/autosplit.py

@@ -0,0 +1,30 @@
+import os
+import random
+import shutil
+
+def split_dataset(frame_folder="dataset/annotated",
+                  train_folder="dataset/train",
+                  val_folder="dataset/val",
+                  split=0.8):
+
+    os.makedirs(train_folder, exist_ok=True)
+    os.makedirs(val_folder, exist_ok=True)
+
+    images = [i for i in os.listdir(frame_folder) if i.endswith(".jpg")]
+    random.shuffle(images)
+
+    split_index = int(len(images) * split)
+    train = images[:split_index]
+    val = images[split_index:]
+
+    for img in train:
+        shutil.copy(os.path.join(frame_folder, img),
+                    os.path.join(train_folder, img))
+
+    for img in val:
+        shutil.copy(os.path.join(frame_folder, img),
+                    os.path.join(val_folder, img))
+
+    print("Dataset Split Complete")
+    print("Train:", len(train))
+    print("Val:", len(val))

datasetforge-0.1.0/datasetforge/dataset_preview.py

@@ -0,0 +1,23 @@
+import cv2
+import os
+import numpy as np
+
+def preview(folder, cols=4):
+    files = os.listdir(folder)[:cols*cols]
+    imgs = []
+
+    for f in files:
+        img = cv2.imread(os.path.join(folder, f))
+        if img is None:
+            continue
+        img = cv2.resize(img, (200,200))
+        imgs.append(img)
+
+    rows = []
+    for i in range(0, len(imgs), cols):
+        rows.append(np.hstack(imgs[i:i+cols]))
+    if rows:
+        grid = np.vstack(rows)
+        cv2.imshow("Preview", grid)
+        cv2.waitKey(0)
+        cv2.destroyAllWindows()

datasetforge-0.1.0/datasetforge/dataset_stats.py

@@ -0,0 +1,23 @@
+import cv2
+import os
+import numpy as np
+
+def dataset_stats(folder):
+    """Show basic stats of dataset"""
+    files = os.listdir(folder)
+    total = 0
+    widths = []
+    heights = []
+
+    for f in files:
+        img = cv2.imread(os.path.join(folder, f))
+        if img is None:
+            continue
+        h, w, _ = img.shape
+        widths.append(w)
+        heights.append(h)
+        total += 1
+
+    print(f"Images: {total}")
+    print(f"Average width: {int(np.mean(widths))}")
+    print(f"Average height: {int(np.mean(heights))}")

datasetforge-0.1.0/datasetforge/dataset_viewer.py

@@ -0,0 +1,24 @@
+import os
+import cv2
+
+def dataset_stats(folder_path):
+    images = [f for f in os.listdir(folder_path) if f.endswith(('.png', '.jpg', '.jpeg'))]
+    print(f"Images: {len(images)}")
+    if not images:
+        return
+    widths, heights = [], []
+    for img_name in images:
+        img = cv2.imread(os.path.join(folder_path, img_name))
+        h, w = img.shape[:2]
+        widths.append(w)
+        heights.append(h)
+    print(f"Average width: {sum(widths)//len(widths)}")
+    print(f"Average height: {sum(heights)//len(heights)}")
+
+def preview_dataset(folder_path, num_images=5):
+    images = [f for f in os.listdir(folder_path) if f.endswith(('.png', '.jpg', '.jpeg'))][:num_images]
+    for img_name in images:
+        img = cv2.imread(os.path.join(folder_path, img_name))
+        cv2.imshow("Preview", img)
+        cv2.waitKey(500)
+    cv2.destroyAllWindows()

datasetforge-0.1.0/datasetforge/object_counter.py

@@ -0,0 +1,9 @@
+import os
+
+def dataset_stats(folder="dataset/annotated"):
+    files = [f for f in os.listdir(folder) if f.endswith(".jpg")]
+    print("\nDATASET STATISTICS")
+    print("-------------------")
+    print(f"Annotated Images: {len(files)}")
+    # You can extend to count objects if needed
+    print(f"Total Objects: 0")

datasetforge-0.1.0/datasetforge/video_dataset.py

@@ -0,0 +1,22 @@
+import cv2
+import os
+
+def video_to_dataset(video_path="dataset/video.mp4", output_folder="dataset/video_frames"):
+    if not os.path.exists(video_path):
+        print("No video found. Put video.mp4 in project folder.")
+        return
+
+    os.makedirs(output_folder, exist_ok=True)
+    cap = cv2.VideoCapture(video_path)
+    count = 0
+
+    while True:
+        ret, frame = cap.read()
+        if not ret:
+            break
+        frame_path = os.path.join(output_folder, f"frame_{count:04d}.jpg")
+        cv2.imwrite(frame_path, frame)
+        count += 1
+
+    cap.release()
+    print(f"Extracted {count} frames to {output_folder}")

datasetforge-0.1.0/datasetforge/webcam_dataset.py

@@ -0,0 +1,35 @@
+# datasetforge/webcam_dataset.py
+import cv2
+import os
+
+def webcam_to_dataset(output_folder="dataset/video_frames"):
+    """
+    Capture frames from webcam and save as images.
+    
+    Args:
+        output_folder (str): Folder to save captured frames.
+    """
+    os.makedirs(output_folder, exist_ok=True)
+    cap = cv2.VideoCapture(0)
+    count = 0
+
+    print("Press 'SPACE' to capture frame, 'q' to quit")
+
+    while True:
+        ret, frame = cap.read()
+        if not ret:
+            break
+        cv2.imshow("Webcam", frame)
+        key = cv2.waitKey(1) & 0xFF
+
+        if key == ord(" "):
+            filename = os.path.join(output_folder, f"frame_{count:04d}.jpg")
+            cv2.imwrite(filename, frame)
+            print(f"Saved: {filename}")
+            count += 1
+        elif key == ord("q"):
+            break
+
+    cap.release()
+    cv2.destroyAllWindows()
+    print(f"Captured {count} frames to {output_folder}")

datasetforge-0.1.0/datasetforge.egg-info/PKG-INFO

@@ -0,0 +1,14 @@
+Metadata-Version: 2.4
+Name: datasetforge
+Version: 0.1.0
+Summary: Computer Vision Dataset Creation Tool
+Author: Shruti Shirpurkar
+Requires-Python: >=3.8
+License-File: LICENSE
+Requires-Dist: opencv-python
+Requires-Dist: numpy
+Dynamic: author
+Dynamic: license-file
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary

datasetforge-0.1.0/datasetforge.egg-info/SOURCES.txt

@@ -0,0 +1,19 @@
+LICENSE
+README.md
+pyproject.toml
+setup.py
+datasetforge/__init__.py
+datasetforge/annotator.py
+datasetforge/auto_annotate.py
+datasetforge/autosplit.py
+datasetforge/dataset_preview.py
+datasetforge/dataset_stats.py
+datasetforge/dataset_viewer.py
+datasetforge/object_counter.py
+datasetforge/video_dataset.py
+datasetforge/webcam_dataset.py
+datasetforge.egg-info/PKG-INFO
+datasetforge.egg-info/SOURCES.txt
+datasetforge.egg-info/dependency_links.txt
+datasetforge.egg-info/requires.txt
+datasetforge.egg-info/top_level.txt

datasetforge-0.1.0/datasetforge.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

datasetforge-0.1.0/datasetforge.egg-info/requires.txt

@@ -0,0 +1,2 @@
+opencv-python
+numpy

datasetforge-0.1.0/datasetforge.egg-info/top_level.txt

@@ -0,0 +1 @@
+datasetforge

datasetforge-0.1.0/pyproject.toml

@@ -0,0 +1,3 @@
+[build-system]
+requires = ["setuptools", "wheel"]
+build-backend = "setuptools.build_meta"

datasetforge-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

datasetforge-0.1.0/setup.py

@@ -0,0 +1,14 @@
+from setuptools import setup, find_packages
+
+setup(
+    name="datasetforge",
+    version="0.1.0",
+    author="Shruti Shirpurkar",
+    description="Computer Vision Dataset Creation Tool",
+    packages=find_packages(),
+    install_requires=[
+        "opencv-python",
+        "numpy"
+    ],
+    python_requires=">=3.8",
+)