Set up and manage DOODS object detection cameras entirely through the Home Assistant UI.
Run two independent detection profiles on the same camera, such as a full-frame and a cropped zone.
Automatically migrate an existing YAML-based DOODS setup into the new UI with nothing lost.
| thebartoman/ha-doods | 0xallam/my-recipe | 0xhassaan/nn-from-scratch | |
|---|---|---|---|
| Stars | 0 | — | 0 |
| Language | Python | Python | Python |
| Last pushed | — | 2022-11-22 | — |
| Maintenance | — | Dormant | — |
| Setup difficulty | easy | moderate | moderate |
| Complexity | 2/5 | 2/5 | 4/5 |
| Audience | general | general | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires a running DOODS detection server to connect to, the integration itself only handles the Home Assistant side.
ha-doods is a modified version of Home Assistant's built in DOODS integration, which connects Home Assistant to a Dedicated Open Object Detection Service to spot things like people or cars in camera images. The original built in version can only be set up by hand editing a configuration file. This fork replaces that with a full setup screen inside Home Assistant itself, under Settings then Devices and Services, so no text file editing is required. Through that setup screen you can connect to your DOODS server, choose which detector it should use, and then add, edit, or remove cameras one at a time. For each camera you can set separate confidence thresholds per label, so different objects can require different certainty levels before triggering, define a specific area of the image to watch, choose where an annotated snapshot image should be saved, and set how often that camera gets checked. It also allows more than one independent detection setup on the very same camera, for example one general profile watching the whole frame and a second one cropped to just a smaller zone, something the original YAML based setup could not do with a single camera entry. If you already had the old style setup in your configuration file, this integration automatically imports it the first time it starts after installing, carrying over every camera, label, confidence level, detection area, snapshot path, and scan interval exactly as they were. Once that import happens, Home Assistant shows a repair notice reminding you that the old configuration text is now redundant and safe to delete, and that notice clears itself automatically once you remove it. Installation is done either through HACS, a popular Home Assistant add-on manager, by adding this project as a custom repository, or manually by copying its folder into Home Assistant's custom components directory. The project notes that it has not yet been submitted back to the official Home Assistant codebase, and it is based on and credits the original built in DOODS integration, which is licensed under the Apache License 2.0.
A fork of Home Assistant's DOODS object detection integration that adds full point and click setup instead of YAML editing.
Mainly Python. The stack also includes Python, Home Assistant, HACS.
Based on code licensed under the Apache License 2.0, which allows free use, modification, and distribution including commercially, as long as the license and notices are kept.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.