From d0581e922e99881c2bef31eb7cbcd685fc76c27f Mon Sep 17 00:00:00 2001 From: rafelafrance Date: Sun, 10 Mar 2024 12:12:27 -0400 Subject: [PATCH] More readme updates --- README.md | 6 +++--- finder/pylib/box_calc.py | 2 +- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index a5e7c97..457c4ad 100644 --- a/README.md +++ b/README.md @@ -27,7 +27,7 @@ cd /path/to/label_finder make install ``` -You will also need to install [YOLO7](https://github.com/WongKinYiu/yolov7). You may follow the instructions given there. I find that if you clone the repository, set up a virtual environment for that repository, and install the requirements into that virtual environment things work just fine. +You will also need to install [YOLO7](https://github.com/WongKinYiu/yolov7). You may follow the instructions given there. I find that if you clone the repository, set up a virtual environment for that repository, and install the requirements into that virtual environment, things work just fine. Every time you want to run any scripts in a new terminal session you will need to activate the virtual environment, once, before running them. @@ -114,9 +114,9 @@ If the sheet is named: `248106.jpg`, then a label may be named `248106_Typewritt ### Optional: Filter typewritten labels -This moves all labels that are classified as "Typewritten" into a separate directory. The OCR works best on typewritten labels or barcodes with printing. It will do a fair job when labels with handwriting or a mix of typewriting & handwriting if they are neatly printed. +This moves all labels that are classified as "Typewritten" into a separate directory. The OCR works best on typewritten labels or barcodes with printing. It will do a fair job with handwritten labels if the handwriting is neatly printed. -I have noticed that the current example YOLO model tends to have a fair number of false positives but close to zero false negatives. Manually pruning the false positives is much easier than sorting all labels. YMMV. +I have noticed that the current example YOLO model (v0.2.0) tends to have a fair number of false positives but close to zero false negatives. Manually pruning the false positives is much easier than sorting all labels. YMMV. #### Example diff --git a/finder/pylib/box_calc.py b/finder/pylib/box_calc.py index e5317e9..fb7c4c4 100644 --- a/finder/pylib/box_calc.py +++ b/finder/pylib/box_calc.py @@ -60,7 +60,7 @@ def find_box_groups(boxes: npt.NDArray, threshold: float = 0.8) -> npt.ArrayLike overlapping[curr] = group found = True # Do we need to look for more overlapping boxes - start = 0 # Used to skip repeated searches + start = 0 # Skip already searched boxes # Every time we find new matches we need to check the new ones against the rest while found: