Iterate on docs

True-North-Intelligent-Algorithms · Feb 16, 2025 · 7cc9436 · 7cc9436
1 parent b4de7ec
commit 7cc9436
Show file tree

Hide file tree

Showing 10 changed files with 56 additions and 23 deletions.
diff --git a/docs/about.md b/docs/about.md
@@ -1,20 +1,16 @@
 # About Napari-Easy-Augment-Batch-dl
 
-Napari-Easy-Augment-Batch-dl is a user-friendly plugin for Napari designed to make batch image augmentation and model training more reproducible . It provides an intuitive graphical interface , integrated with Napari, that allows users to load and label images, apply augmentations, and process large datasets efficiently without requiring programming knowledge.  While the plugin complements programming by streamlining workflows, I always encourage users to learn programming, as it provides deeper flexibility and control over their data and models.
+Napari-Easy-Augment-Batch-dl is a user-friendly plugin for Napari designed to make labeling, batch image augmentation and model training more reproducible . It provides an intuitive graphical interface integrated with Napari, that allows users to load and label images, apply augmentations, and process large datasets efficiently without requiring programming knowledge.  (While the plugin complements programming by streamlining workflows, I always encourage users to learn programming, as it provides deeper flexibility and control over their data and models).
 
 ## 🔍 Features
-- Takes advantage of Napari's build in labelling functionality Easy-to-use GUI .
+- Takes advantage of Napari's built in labelling functionality.
 - Supports various augmentation techniques such as rotation, flipping, noise addition, and brightness adjustments.
-- Explicitly saves augmented patches for easier trouble shooting and more repeatible training. 
-- Streamlined workflow with three simple steps: Load & Label, Configure Augmentations, and Train & Predict.
-- Plugin mechanism to support training 
-- Ideal for researchers, data scientists, and image analysis professionals.
-
-## 🎯 Purpose
-The primary goal of Napari-Easy-Augment-Batch-dl is to bridge the gap between advanced image augmentation techniques and users who need a simple, no-code solution. Whether you are working with machine learning datasets, medical imaging, or scientific research, this tool helps you prepare high-quality augmented images efficiently.
+- Explicitly saves augmented patches for easier trouble shooting and more repeatable training.  
+- Streamlined workflow with three simple steps: Load & Label, Augmentat, and Train & Predict.
+- Plugin mechanism to support training and prediction with different frameworks (Cellpose, Stardist, SAM, Yolo, UNET, and more)
 
 ## 🛠️ Development & Contributions
-Napari-Easy-Augment-Batch-dl is an open-source project, and contributions are welcome! If you have feature requests, bug reports, or would like to contribute, visit our [GitHub repository](https://github.com/YOUR_REPO).
+Napari-Easy-Augment-Batch-dl is an open-source project, and contributions are welcome! If you have feature requests, bug reports, or would like to contribute, visit our [GitHub repository](https://github.com/True-North-Intelligent-Algorithms/napari-easy-augment-batch-dl).
 
 ## 📞 Support
-For questions, issues, or feedback, please check our [FAQ](faq.md) or open a discussion on our GitHub page.
+For questions, issues, or feedback, please check our [FAQ](faq.md) or open a discussion on [image.sc](image.sc).
diff --git a/docs/augment.md b/docs/augment.md
@@ -1,14 +1,23 @@
 # Configure Augmentations  
 
-Set up your augmentation settings before processing your images.  
+You need to generate augmentations before training. 
 
 ## 🎛️ Augmentation Panel  
 ![Augmentation Panel](images/augment_panel.png)  
 
 ### Steps:  
-1️⃣ Select the augmentation types (rotation, flipping, noise, etc.).  
-2️⃣ Adjust the parameters as needed.  
-3️⃣ Preview the augmentations before applying.  
+1️⃣ Make sure you draw at least one label box and label objects within it. See [here](https://true-north-intelligent-algorithms.github.io/napari-easy-augment-batch-dl/load_and_label/)  
+2️⃣ Select the augmentation types (rotation, flipping, brightness/contrast, etc.)  
+3️⃣ Choose ```Augment current image``` or ```Augment all images```  
+4️⃣ Verify by looking in the ```patches``` sub-directory of your project.
+
+After completing above steps you should find a ```ground truth``` and ```input``` directory has been created in your project in the ```patches directory```  
+
+![Patches top level](images/patches_top_level.jpg)
+
+Within the ```input``` directory you should see a collection of patches from your images, in the ```ground truth``` directory there should be corresponding patches taken from the label image.  Make sure you draw at least one label box and label objects within it. See (here)[]
+
+![Patches](images/patches.jpg)
 
 ---
 

diff --git a/docs/faq.md b/docs/faq.md
@@ -11,6 +11,7 @@ This tool is designed for users who need to augment images in bulk, such as rese
 ## 🔄 Usage Questions
 
 ### How do I install the plugin?
-You can install it using pip:
+You can install it using pip:  
 ```sh
-pip install napari-easy-augment-batch-dl
+pip install napari-easy-augment-batch-dl
+```
diff --git a/docs/images/labels_dir.png b/docs/images/labels_dir.png
diff --git a/docs/images/patches.jpg b/docs/images/patches.jpg
diff --git a/docs/images/patches_top_level.jpg b/docs/images/patches_top_level.jpg
diff --git a/docs/images/project_dir_save.png b/docs/images/project_dir_save.png
diff --git a/docs/images/project_dir_start.png b/docs/images/project_dir_start.png
diff --git a/docs/load_and_label.md b/docs/load_and_label.md
@@ -1,19 +1,41 @@
 # Load and Label
 
-The first step is to load your images and assign labels.
+## Preparation
+
+Prior to startingwith this plugin, put the images you want to work with in a project directory.  
+
+![Project Directory](images/project_dir_start.png)
 
 ## 📌 Load Panel  
+
+After starting the plugin the first step is to load your images and assign labels.  
+
 ![Load Panel](images/load_panel.png)  
 
-### Steps:  
 1️⃣ Click the **Open image directory...** button.  
 2️⃣ Select the directory that contains your image files.  
-3️⃣ Assign labels as needed:
-   - a) Select **Label box** layer and draw a label box that is as large or larger than the desired patch size.
-   - b) Select **labels** layer and Label objects within the label box.
+
+## Drawing Labels
+
+1️⃣ Select **Label box** layer and draw a label box that is as large or larger than the desired patch size.  
+2️⃣ Select **labels** layer and Label objects within the label box.
 
 ![Label Box and Labels](images/label_box_and_labels.png)
 
+
+## Save Results
+
+Select ```Save Results``` periodically to save the labels you have drawn.  
+
+![Load Panel](images/load_panel.png)  
+
+After saving results folders should be generated for different types of deep learning artifacts.  
+
+![Project Directory](images/project_dir_save.png)
+
+Inspect the labels directory to verify labels you have drawn have been saved.  
+
+![Labels Directory](images/labels_dir.png)
 ---
 
 🔄 **Next:** [Configure Augmentations](configure_augmentations.md)
diff --git a/docs/train_and_predict.md b/docs/train_and_predict.md
@@ -6,10 +6,15 @@ Train a model using your labeled data and make predictions on new images.
 ![Training Panel](images/train_panel.png)  
 
 ### Steps:  
-1️⃣ Choose a model or configure training parameters.  
+1️⃣ Choose a model from dropdown and configure training parameters.  
 2️⃣ Train using labeled images.  
 3️⃣ Use the trained model to predict labels on new images.  
 
+## Training popup 
+
+After hitting train a popup should appear which allows you to further adjust training parameters.  
+
+![Training popup](images/train_cellpose.png)
 ---
 
 🔄 **Next:** [Run & Export](run_and_export.md)