IT – DK FILMS

IT PROJECTS

Censor_video-photo_app

customer-relationship-management-concept

The application is designed for automatic censorship of images and videos, as well as batch processing of folders with files. It analyzes content, hides unwanted elements, and saves the results with the same structure as the original data.

GitHub

Key Features

Photo and video censorship – Detects and processes content by automatically hiding unwanted areas.
Folder processing – Analyzes the input directory and creates an output folder with the same structure, but with edited files.
Automatic scanning – The application finds and processes all images and videos in the specified directory.
Supports various formats – Works with popular video formats (MP4, AVI, MOV) and image formats (JPG, PNG).
High performance – Uses powerful Nudenet-based models for accurate detection and processing.
Customizable labels and threshold settings – Users can adjust the detection threshold and select specific labels for processing.
Three censorship modes – Users can choose:

🔵 Blur

🔲 Pixelation
◼️ Black rectangles

Who is this app for?

Owners of platforms with user-generated content.
Content moderators.
Personal users who need automatic media censorship.

How does it work?

Select a folder with images and videos (optionally, specify the output directory, or let the program create one automatically).
Choose the censorship mode, adjust labels (in addition to the default ones), and set the detection threshold (by default, the highest threshold is applied).
The application analyzes and processes the files.
The final result is saved with the same structure as the original files.

How does it work?

Select a folder with images and videos (optionally, specify the output directory, or let the program create one automatically).
Choose the censorship mode, adjust labels (in addition to the default ones), and set the detection threshold (by default, the highest threshold is applied).
The application analyzes and processes the files.
The final result is saved with the same structure as the original files.

Technologies Used

Nudenet (NudeDetector) – Detecting unwanted content.
Pillow (PIL) – Image processing, blurring, pixelation.
OpenCV (cv2) – Video processing.
FFmpeg – Video file encoding and decoding.
PyQt5 – Graphical user interface.
NumPy – Image array operations.
subprocess – Command-line execution (FFmpeg processing).
shutil, os – File and directory management.

In Development

Enhancing the detection model for improved accuracy.
Built-in optimization functions for better detection quality.
More precise control over censorship types and additional customization options.