Changelog¶

v1.2.0 (01/03/2025)¶

GSV Metadata Update Capability: Added new update_metadata method for GSVDownloader to update year/month metadata for existing panorama IDs without redoing the entire download process.

KartaView API Fix: Fixed KartaView download API and added retry decorator for improved reliability.
Object Detection Enhancement: Added new object detection functionality.
Improved Segmentation: Enhanced pixel ratio calculation in segmentation with improved normalization.
Path Handling Improvement: Using Path() for path_pid existence check for better cross-platform compatibility.
Dependency Management: Updated poetry.lock with new dependencies and version upgrades.
Removed Deprecated Modules: Cleaned up codebase by removing deprecated modules.
Added Example Notebooks: Included Jupyter notebooks demonstrating download, metadata, transform, and visualization capabilities.
Verbosity Control: Added enhanced verbosity control for better user experience.
Documentation Update: Updated README.md with improved documentation.

License Changed: Changed license from CC BY-SA 4.0 to MIT License
AMSDownloader Introduced: New downloader for Amsterdam Street View Images (SVI). Usage: from zensvi.download import AMSDownloader.
PointCloudProcessor Added: New module to convert Street View Images to point clouds. Usage: from zensvi.transform import PointCloudProcessor.
Embeddings Module Added: New functionality for generating image embeddings. Usage: from zensvi.cv import Embeddings.
Enhanced Documentation: Improved documentation for embeddings functionality.
Enhanced Metadata Return: Improved metadata return functionality in MLYDownloader.
Removed Geopolars Dependency: Streamlined dependencies by removing geopolars requirement.

ClassifierPerception Introduced: New classifier for perception-based image analysis. Usage: from zensvi.cv import ClassifierPerception.

Metadata Module Added: New module for metadata processing. Usage: from zensvi import metadata.
ClassifierGlare, ClassifierLighting, ClassifierPanorama, ClassifierPlatform, ClassifierQuality, ClassifierReflection, ClassifierViewDirection, ClassifierWeather Introduced: New classifiers for various image quality attributes. Usage: from zensvi.cv import ClassifierGlare, ClassifierLighting, ClassifierPanorama, ClassifierPlatform, ClassifierQuality, ClassifierReflection, ClassifierViewDirection, ClassifierWeather.
DepthEstimator Introduced: New module for depth estimation. Usage: from zensvi.cv import DepthEstimator.

Visualization Module Added: Introducing a new module for plotting maps and images. Usage: from zensvi import visualization.
Low Level Features in CV Module: New function to quantify low-level features. Usage: from zensvi.cv import get_low_level_features.
ClassifierPlaces365 Introduced: New classifier based on Places365 model. Usage: from zensvi.cv import ClassifierPlaces365.

Improved Download Efficiency: MLYDownloader now uses polygons for improved download efficiency. GSVDownloader improved data cleaning to eliminate overlaps and dark spots.
Metadata-Only Downloads: Added metadata_only argument to download only the metadata of SVIs.

Flexible Image Transformation: Introduced theta argument in ImageTransformer to modify view angles without affecting the FOV.

Sub-Folder Creation in Downloads: Added batch_size argument to organize downloaded SVIs into sub-folders, addressing performance issues with large image counts.

Panoptic Segmentation Task: Enabled panoptic segmentation along with semantic segmentation.
Flexible Output Formats: Introduced flexible csv formats and multiple output options including pixel ratios.

Enhanced Download SVI Arguments: New arguments for cropping, saving full images or tiles, and OSMnx enhancements.

Enhanced Image Transformation Options: Added FOV and aspects arguments, and introduced new fisheye transformation methods.

New Downloaders for Street Views: Deprecated older downloader in favor of new GSVDownloader and MLYDownloader.

New Argument in download_gsv Function: input_place_name for specifying the boundary for downloading GSV.

First release of zensvi!
Improved Spatial Processing: Enhanced speed and introduced ID columns for better tracking of input and panorama IDs.