Efficient Fruit Defect Detection and Glare removal Algorithm by anisotropic diffusion and 2D Gabor filter
This is an incremental improvement for agricultural quality control, specifically targeting fruit inspection under uneven lighting conditions.
The paper tackles fruit defect detection by addressing glare removal as a preprocessing step using anisotropic diffusion and morphological operations, and then applies 2D Gabor filters for segmentation, resulting in enhanced defect retrieval and robustness.
This paper focuses on fruit defect detection and glare removal using morphological operations, Glare removal can be considered as an important preprocessing step as uneven lighting may introduce it in images, which hamper the results produced through segmentation by Gabor filters .The problem of glare in images is very pronounced sometimes due to the unusual reflectance from the camera sensor or stray light entering, this method counteracts this problem and makes the defect detection much more pronounced. Anisotropic diffusion is used for further smoothening of the images and removing the high energy regions in an image for better defect detection and makes the defects more retrievable. Our algorithm is robust and scalable the employability of a particular mask for glare removal has been checked and proved useful for counteracting.this problem, anisotropic diffusion further enhances the defects with its use further Optimal Gabor filter at various orientations is used for defect detection.