Image Processing of Metal Parts using Quality Inspection Method in Comparison with MATLAB

this paper proposes a novel image processing approach for rapid quality inspection in metal parts of machinery products. The quality inspection approaches have evolved from the traditional visual inspection to sophisticated techniques like automated pattern recognition. The paper concentrates a novel image processing algorithm for rapid and cost-effective quality detection of metal surfaces. The process deals with analysis of the metal product images and extracting the crack location using appropriate thresholding, Hence, these values are determined from the histogram of the mental image from which the deviations between weld surface and normal surface being found from ranging intensity values. The images of the metal products are acquired with the help of calibrated camera sensors in an ideal lab set up in order to avoid the illumination effects which will interfere with the crack surface. The mental image enhances precursor for crack-feature detection. The cracks in the metal products are detected analyzed using the MATLAB interface where the Image Processing Toolbox provides a comprehensive set of reference-standard algorithms and functions for analysis, Visualization and algorithm development. The dimensional analysis and qualitative aspects of the crack are recorded to determine the quality status of the metal product. Thus, this approach is implemented for four mild steel specimens and a higher accuracy in the detection process is observed. Further, the compatibility of the algorithm with metal surface with varied sizes and colors are analyzed.