Retrospective Shading Correction of Confocal Laser Scanning Microscopy Beef Images for Three-Dimensional Visualization

Because of the inherent imperfections of the image formation process, confocal laser scanning microscopy (CLSM) images are often corrupted by spurious intensity variations not present in the original scene, which is usually referred to as shading or intensity inhomogeneity. In this paper, a retrospective shading correction method for CLSM beef images was developed using a hybrid of image processing algorithms. A partial-differential-equation-based diffusion technique was applied, firstly, to reduce the additive shading component. To reduce the computational burden, a thresholding segmentation method was then presented to separate the muscle tissues from the background. After that, a robust, automatic, and fast method was applied for reduction of multiplicative shading component. The corrected images were finally used to construct a three-dimensional beef image, which could provide a valuable source of knowledge about beef microstructure that cannot be obtained via traditional two-dimensional images.