A simple windows application to search for images in a directory.
- Search for text in images using OCR. Tested with English, Traditional Chinese, Simplified Chinese.
- Search for objects in images using YOLOv8. Labels from COCO or Open Image V7.
- Search for images using its class. Labels from ImageNet.
- Supported image formats: formats supported by OpenCV: bmp, dib, jpeg, jpg, jpe, jp2, png, webp, avif, pbm, pgm, ppm, pxm, pnm, pfm, sr, ras, tiff, tif, exr, hdr, pic.
- Download the latest release
- Run the application
- Select the directory to Index using models (only required once)
- Search
If you clone the repository:
- Install the required packages using Poetry.
- Put the ONNX format YOLOv8 models in the
models
directory.
The first time you run the application, it will take some time to index the images in the directory.
Only YOLOv8n and YOLOv8n COCO models are included in the minimal release. For more models, download the ONNX format models and put them in the models
directory.
-
Using cx_Freeze
python dist.py build
-
Using Nuitka
python -m nuitka main.py --standalone --enable-plugin=pyside6 --include-data-dir=./backend/libsimple-windows-x64=lib/backend/libsimple-windows-x64 --include-data-files=./backend/libsimple-windows-x64/simple.dll=lib/backend/libsimple-windows-x64/simple.dll --include-data-dir=./models=models --include-data-dir=./.venv/Lib/site-packages/rapidocr_onnxruntime=rapidocr_onnxruntime --disable-console
- OCR is done using RapidOCR
- Object detection adapted from ONNX-YOLOv8-Object-Detection
- Search tokenizer from Simple
- Object detection and image classification model from YOLOv8