🧠 A Complete Guide to ML AI Image Classification

PhenoCapture’s ML Image Classification is a state-of-the-art, end-to-end Machine Learning suite built directly into your workflow. Whether you want to count specific types of cells, categorize manufacturing defects, or sort complex biological structures, this tool empowers you to train your own Artificial Intelligence to do the heavy lifting.

Here is everything you need to know to extract data, train a Deep Neural Network, and automatically classify your images.

💡 Key Features

• Smart Data Extraction: Easily generate a massive training dataset from a single image. Using advanced AI Binarization and smart screening conditions (like area and width minimums), the tool automatically isolates individual objects and saves them as neatly cropped images.
• 1-Click Training Presets: You don’t need a degree in data science! Choose from pre-configured training presets—ranging from Ultra-Fast (Pipeline Test) for quick drafts to Maximum Accuracy for deep learning—and let the software automatically configure the complex settings for you.
• Visual Classification Overlays: Once your AI is trained, running an analysis places beautiful, color-coded outlines around your objects. You can easily assign specific colors to different classification categories for instant visual verification.
• High-Speed Batch Processing: Have hundreds of images to classify? Load them into the Batch Processing engine, hit Start, and the AI will analyze every single file sequentially, completely automating your workflow.
• Comprehensive Data Export: Every analysis automatically generates detailed visual overlays, individual cropped masks, and a complete CSV spreadsheet containing the exact coordinates, pixel areas, and predicted labels for every detected object.

🛠️ How to Use ML Image Classification

Training your AI and running an analysis is broken down into three logical phases:

Phase 1: Prepare Your Training Data

First, you need to teach the AI what to look for.

1. Load a source image and adjust your Binarization and Screening Conditions to perfectly isolate your target objects.
2. Click Create Training Images. The software will crop out every individual object and save them. Sort these images into their correct category folders (e.g., Folder 0 for Class A, Folder 1 for Class B) to create your dataset.

Phase 2: Train Your Model

Once your dataset is sorted:

1. Select your main folder containing the categorized images.
2. Choose a Training Preset from the dropdown menu (e.g., Normal (Balanced)).
3. Click Begin Training. The software will train your custom AI model and generate a TrainedModel.zip file. You can even click Begin Validation to test its accuracy and see a detailed report!

Phase 3: Analyze and Classify

Now it is time to put your AI to work on new images!

1. Select your trained model from the Model List.
2. Open a new image (or load an entire folder in the Batch Processing tab).
3. Adjust your View Options to assign unique outline colors to your different AI classes.
4. Click Perform Analysis. The AI will instantly detect, classify, and color-code every object, generating a ready-to-use CSV report alongside your stunning visual results!

With the ML Image Classification tool, building and deploying custom, highly accurate Artificial Intelligence is incredibly accessible and completely automated!