<aside> 🛠

Crop Images with Precise Coordinate Boundaries

Crop images using explicit bounding box coordinates (top, left, bottom, right). Specify exact pixel boundaries to extract rectangular regions. Perfect for precise cropping, automation workflows, programmatic cropping, or when you have coordinates from external systems. Use with detection/segmentation outputs or when you know exact coordinates

</aside>

What it does

Image Crop by Coord extracts a rectangular region from an image using explicit bounding box coordinates. Specify the region boundaries: top, left, bottom, right (in pixels). The tool extracts only that rectangular region. No detection or guessing—you provide exact coordinates, it crops exactly. Perfect for automation, programmatic workflows, or when coordinates come from another system.

Problem it solves

• Precise rectangular cropping – Crop exact rectangular regions by coordinates
• Automation & API workflows – Use with coordinates from detection systems
• Programmatic cropping – Integrate coordinates from other tools/systems
• Batch cropping – Apply same coordinates to multiple images
• Detection output processing – Use bounding boxes from detection models
• Coordinate-based extraction – Extract regions when you have coordinates
• Region validation – Test specific coordinate regions
• Multiple region extraction – Crop multiple regions from one image

Input/Output

<aside>

• Input: Image + Bounding Box Coordinates
  • Image Format: Any image format
  • Coordinates: Top, Left, Bottom, Right (in pixels)
    • Top: Y-coordinate of top edge (pixels from top)
    • Left: X-coordinate of left edge (pixels from left)
    • Bottom: Y-coordinate of bottom edge (pixels from top)
    • Right: X-coordinate of right edge (pixels from left)

• Output: Cropped region image
  • Format: Cropped image
  • Dimensions: (Bottom - Top) × (Right - Left) pixel </aside>

How to Use

<aside>

Crood.mp4

1. Load detection → Get output from object detection model (DETR, YOLOS, etc.)
2. Select target object to crop → Choose target object and check its coordinates
3. Set coordinates → Upload input text box, enter Top, Left, Bottom, Right values (in pixels), and ‘Add Text’
4. Run Flow </aside>