Inaccurate Object Mask Alignment with Image Border in yolov5x-segment

Search before asking

[X] I have searched the YOLOv5 issues and discussions and found no similar questions.

Question

Issue Description: I have been using yolov5x-segment for my project, and I have successfully annotated my dataset with accurate object masks. However, during testing, I noticed that the object masks for objects located at the image borders do not align perfectly with the image edges. This misalignment results in visible gaps when I later stitch the segmented images together.

Question: I would like to understand if this issue is related to the prediction capability of the yolov5x-segment model. What could be causing this misalignment, and how can I address it to achieve seamless alignment between the object masks and the image borders?

Steps to Reproduce:

Train yolov5x-segment model using annotated dataset.
Perform inference on test images.
Observe misalignment between object masks and image borders.
Notice visible gaps when stitching segmented images together (This is the case even when using sliding window stitching with overlapping pixels).

Expected Behavior: I expected the object masks to align accurately with the image borders, ensuring seamless stitching of segmented images without any visible gaps. split_image_with_predicted_masks after_stitching

Additional

No response

👋 Hello @lugushaoye, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution.

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Requirements

Python>=3.7.0 with all requirements.txt installed including PyTorch>=1.7. To get started:

git clone https://github.com/ultralytics/yolov5  # clone
cd yolov5
pip install -r requirements.txt  # install

Environments

YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

Notebooks with free GPU:
Google Cloud Deep Learning VM. See GCP Quickstart Guide
Amazon Deep Learning AMI. See AWS Quickstart Guide
Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training, validation, inference, export and benchmarks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

Introducing YOLOv8 🚀

We're excited to announce the launch of our latest state-of-the-art (SOTA) object detection model for 2023 - YOLOv8 🚀!

Designed to be fast, accurate, and easy to use, YOLOv8 is an ideal choice for a wide range of object detection, image segmentation and image classification tasks. With YOLOv8, you'll be able to quickly and accurately detect objects in real-time, streamline your workflows, and achieve new levels of accuracy in your projects.

Check out our YOLOv8 Docs for details and get started with:

pip install ultralytics

@lugushaoye hi there,

Thank you for reaching out and providing detailed information about the issue you are facing with yolov5x-segment.

The misalignment you are observing between object masks and image borders is not related to the prediction capability of the yolov5x-segment model. It is actually a limitation of the current implementation.

This misalignment occurs because the model predicts object masks that may extend beyond the actual boundaries of the objects. When stitching the segmented images together, this can result in visible gaps.

To address this issue and achieve a seamless alignment between your object masks and image borders, you can try the following approaches:

Increase the image size: By increasing the size of your input images during inference, you can provide more context to the model, which may help it predict object masks that align better with the image edges.
Refine the object masks: After obtaining the predicted object masks, you can apply post-processing techniques to refine them. For example, you can use morphological operations like dilation and erosion to expand or shrink the predicted masks and align them better with the image borders.
Apply padding: Another approach is to apply padding to your images before feeding them to the model. By adding a border around the images, you can ensure that all objects have some space between their boundaries and the image edges, reducing the chances of misalignment.

Please note that these are general suggestions, and the effectiveness of each approach may vary depending on your specific use case. It is recommended to experiment with different techniques and parameters to find the best solution for your application.

I hope this information helps you address the issue. If you have any further questions or concerns, please feel free to ask.

Best regards,

👋 Hello there! We wanted to give you a friendly reminder that this issue has not had any recent activity and may be closed soon, but don't worry - you can always reopen it if needed. If you still have any questions or concerns, please feel free to let us know how we can help.

For additional resources and information, please see the links below:

Docs: https://docs.ultralytics.com
HUB: https://hub.ultralytics.com
Community: https://community.ultralytics.com

Feel free to inform us of any other issues you discover or feature requests that come to mind in the future. Pull Requests (PRs) are also always welcomed!

Thank you for your contributions to YOLO 🚀 and Vision AI ⭐

ultralytics / yolov5