Fix in BBox clipping

[ternaus]

Fixes: #2309

Summary by Sourcery

Bug Fixes:

• Fix issues with bounding box clipping when using the "yolo", "coco", or "pascal_voc" formats.

[sourcery-ai]

Reviewer's Guide by Sourcery

This pull request addresses a bug in bounding box clipping where class IDs were being incorrectly modified when clipping was enabled. The fix ensures that bounding boxes are clipped to the image boundaries without altering the associated class labels or scores. Additionally, comprehensive tests have been added to cover various bounding box formats, edge cases, and data types.

Sequence diagram for BBox clipping process

sequenceDiagram
    participant C as Client
    participant BU as BBoxUtils
    participant LF as LabelField

C->>BU: check_and_convert(bbox_data)
BU->>BU: convert_bbox_format()
BU->>BU: clip_bbox_coordinates()
Note over BU: Only clip x,y coordinates
Note over BU: Preserve class IDs/scores
BU->>BU: filter_invalid_bboxes()
BU-->>C: return processed_bboxes

Loading

Class diagram for BBox and Label Field handling

classDiagram
    class BBoxUtils {
        +params
        +check_and_convert(data)
        -clip_bbox_coordinates()
        -filter_invalid_bboxes()
    }

    class LabelFieldEncoder {
        +is_numerical_label: dict
        +label_encoders: dict
        +_encode_label_field(data, data_name, label_field)
        +_validate_label_field_length(data, data_name, label_field)
    }

    note for BBoxUtils "Modified clipping to preserve class IDs"
    note for LabelFieldEncoder "Enhanced numerical type handling"

Loading

File-Level Changes

Change	Details	Files
Added comprehensive test cases for bounding box clipping with various formats, edge cases, and data types.	Added parameterized tests for 'yolo', 'coco', and 'pascal_voc' formats. Included tests for empty bounding box arrays. Added tests for single bounding boxes with high class IDs. Added tests for edge cases such as very small and very large bounding boxes. Included tests for different numpy data types for class labels and scores. Added a specific test case to reproduce the class ID clipping bug. Added tests for very high and very low confidence scores. Added tests for multiple label fields with different numpy dtypes.	tests/test_bbox.py
Fixed the bounding box clipping logic to prevent modification of class IDs.	Modified the clipping logic to only clip the bounding box coordinates and not the class IDs or scores. Added a check to ensure clipping is only applied if there are bounding boxes.	albumentations/core/bbox_utils.py
Modified the label encoding logic to handle numpy arrays and preserve data types.	Modified the _encode_label_field method to handle numpy arrays directly. Preserve the numpy array dtype when encoding numerical label fields. Added a check to see if the input is a numpy array before converting to float32.	albumentations/core/utils.py

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!
• Generate a plan of action for an issue: Comment @sourcery-ai plan on
  an issue to generate a plan of action for it.

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[sourcery-ai]

Hey @ternaus - I've reviewed your changes and they look great!

Here's what I looked at during the review

• 🟢 General issues: all looks good
• 🟢 Security: all looks good
• 🟡 Testing: 1 issue found
• 🟢 Complexity: all looks good
• 🟢 Documentation: all looks good

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 89.83%. Comparing base (b1a79c2) to head (3c633d8).
Report is 394 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##           main    #2310       +/-   ##
=========================================
+ Coverage      0   89.83%   +89.83%     
=========================================
  Files         0       50       +50     
  Lines         0     9017     +9017     
=========================================
+ Hits          0     8100     +8100     
- Misses        0      917      +917     

Flag	Coverage Δ
ubuntu-py3.12	89.83% <100.00%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.