Deep Learning-Driven of Turbidity Levels Dataset

The dataset is designed for training and evaluating deep learning models to classify water samples based on their turbidity levels. Turbidity, measured in NTU (Nephelometric Turbidity Units), indicates water clarity and quality. The dataset comprises images of water samples categorized into five distinct classes, each representing a different range of NTU values.

Classes:

Class 1: 200-320 NTU

This class represents water with relatively low turbidity within the given range, indicating some level of clarity.

Class 2: 320-440 NTU

This class includes water samples with moderate turbidity, where water has noticeable suspended particles.

Class 3: 440-560 NTU

This class signifies higher turbidity, where water is considerably less clear due to an increased presence of suspended particles.

Class 4: 560-680 NTU

This class represents high turbidity levels, indicating significantly reduced water clarity with many suspended particles.

Class 5: 680-800 NTU

This class includes water samples with very high turbidity, where water is highly opaque due to a large amount of suspended matter.

Image Samples:

The dataset contains images of water samples taken under controlled lighting conditions to ensure consistency.

Each image captures the visual characteristics of the water sample in a transparent container.

Images are labeled with their corresponding NTU values and class.

Data Format:

Images: Stored in NEF format, with standardized resolution and dimensions.

Labels: Accompanied by a .xlsx file containing image filenames and their corresponding NTU values and class labels.

Applications:

The dataset can be used to develop and evaluate deep learning models for automatic classification of water turbidity levels.

Potential applications include environmental monitoring, water treatment facilities, and public health assessments.