An Image Preprocessing Model of Coal and Gangue in High Dust and Low Light Conditions Based on the Joint Enhancement...

Area of Science:

• Mining Engineering
• Image Processing
• Computer Vision

Background:

• Coal mine dust pollution severely degrades image quality, causing dimness, shadows, and reflections.
• This poor image quality hinders accurate identification of coal and gangue from the background.
• Existing methods struggle to address the unique challenges posed by these low-quality images.

Purpose of the Study:

• To develop an effective preprocessing model for low-quality coal and gangue images.
• To enhance image brightness, contrast, and details despite harsh mining conditions.
• To improve the accuracy of coal and gangue identification and subsequent target recognition.

Main Methods:

• Analysis of coal and gangue image characteristics to identify improvement strategies.
• Development of an image preprocessing workflow based on local image features.
• Proposal of a joint image enhancement algorithm utilizing bilateral filtering.

Main Results:

• The proposed model effectively reduces noise in images.
• Overall image brightness and contrast are significantly improved.
• Image details are enhanced, leading to better segmentation results compared to other methods.
• Experimental validation using K-means clustering and metrics like information entropy and structural similarity.

Conclusions:

• The developed preprocessing model provides a robust solution for enhancing low-quality coal and gangue images.
• Improved image quality facilitates more accurate segmentation and target recognition in mining environments.
• This work offers a valuable foundation for automated coal and gangue identification systems.