Mask Detection

This function block is designed to analyze images and detect whether individuals are wearing masks properly. It utilizes a weight-based detection algorithm to effectively categorize the usage of masks within the input image.

📥 Inputs

Image RGB An RGB image that contains persons for mask detection.

📤 Outputs

Image RGB The output will be a processed image displaying detection results.

Masked The count of correctly worn masks.

Uncorrect Masked The count of incorrectly worn masks.

No Mask The count of individuals not wearing masks.

🕹️ Controls

Confidence Threshold % A slider to adjust the required confidence level for mask detection. Values can be set between 40% to 90%.

🎨 Features

Real-time Mask Detection The function block can efficiently analyze images in real-time and provide feedback on mask usage.

Visual Output The output image displays detection results, helping users quickly understand mask compliance.

📝 Usage Instructions

1. Input Image: Connect an RGB image that contains people who may or may not be wearing masks to the Image RGB input.

2. Adjust Confidence Level: Use the Confidence Threshold % slider to set how strict the detection should be.

3. Run Analysis: Execute the block to analyze the input image and receive an output with detection results.

📊 Evaluation

The function block will analyze the input image, running the mask detection algorithm and returning the updated image along with counts of different mask conditions.

💡 Tips and Tricks

Improving Detection Accuracy

Increasing the confidence threshold may yield more accurate results, but be aware that it might also lead to fewer detections. Test different settings to find the optimal balance.

Batch Processing

For processing multiple images, consider using the Batch Processing block after configuring this function block. It will help reduce memory usage during large-scale detections.

🛠️ Troubleshooting

No Detections Found

If no detections are made, ensure that the input image is clear and contains faces that are mostly frontal. Low-quality or blurry images may hinder detection.

Incorrect Mask Counting

In cases where you find discrepancies in counting the mask usage, adjust the Confidence Threshold % to improve detection quality or consider re-evaluating the input image quality.