An Integrated Horizon Picking Method for Obtaining the Main and Detailed Reflectors on Sub-bottom Profiler Sonar Image

A sub-bottom profiler (SBP) can capture the sediment interfaces and properties of different types of sediment. Horizon picking from SBP images is one of the most crucial steps in marine sub-bottom sediment interpretation. However, traditional horizon picking methods are good at obtaining the main horizons representing the main reflectors while ignoring the detailed horizons. While detailed horizons are the prime objective, many tiny structures caused by interference echoes will also be picked. To overcome this limitation, an integrated horizon picking method for obtaining the main and detailed horizons simultaneously is proposed in this paper. A total of three main process steps: the diffusion filtering method, the enhancement filtering method as well as the local phase calculation method, are used to help obtain the main and detailed horizons. The diffusion filtering method smooths the SBP images and preserves reflectors. Enhancement filtering can eliminate outliers and enhance reflectors. The local phase can be used to highlight all of the reflections and help in the choosing of detailed horizons. A series of experiments were then performed to validate the effectiveness of the proposed method, and good performances were achieved.

Defect identification in magnetic tile images using an improved nonlinear diffusion method

Visual inspection of surface defects is a crucial step in the magnetic tile manufacturing process. Magnetic tile images suffer from a non-uniform illumination, texture and noise that disperse irregularly in flawless image areas. As a result, common edge detection and threshold segmentation techniques fail to identify these kinds of defects. In this work, we present a robust algorithm for defect identification in magnetic tile images. The proposed method is based on a new anisotropic diffusion filtering model. Unlike traditional anisotropic diffusion models that take into account only gradient magnitude information, the proposed model combines together gradient magnitude and a new local difference image feature. The aim is to remove bright shapes and undesirable artifacts in the faultless region in magnetic tile images. In addition, the method activates a smoothing process in the flawless region to homogenize the background and simultaneously a sharpening in the defect boundaries to highlight anomalies. Experimental results on a number of magnetic tiles samples containing different types of defects have demonstrated the efficiency of the proposed diffusion method.

Approach for semi-automated measurement of fiber diameter in murine and canine skeletal muscle

Currently available software tools for automated segmentation and analysis of muscle cross-section images often perform poorly in cases of weak or non-uniform staining conditions. To address these issues, our group has developed the MyoSAT (Myofiber Segmentation and Analysis Tool) image-processing pipeline. MyoSAT combines several unconventional approaches including advanced background leveling, Perona-Malik anisotropic diffusion filtering, and Steger’s line detection algorithm to aid in pre-processing and enhancement of the muscle image. Final segmentation is based upon marker-based watershed segmentation. Validation tests using collagen V labeled murine and canine muscle tissue demonstrate that MyoSAT can determine mean muscle fiber diameter with an average accuracy of ~92.4%. The software has been tested to work on full muscle cross-sections and works well even under non-optimal staining conditions. The MyoSAT software tool has been implemented as a macro for the freely available ImageJ software platform. This new segmentation tool allows scientists to efficiently analyze large muscle cross-sections for use in research studies and diagnostics.