Wavelet Transform In The Tensile Fracture Of The Composite Image Detection

In this paper, a special wavelet transform used to detect snapped exterior of the fabric composite is proposed. Generally speaking, the regularity of orthonormal wavelet bases with compact support increases linearly with the support width. But if we relax the orthogonality, much more freedom on the choice of wavelet function is gained. Based on this idea, in this paper, we used an infinitely differentiable wavelet with a local support. Its properties are different from those of orthonormal wavelet bases. It can be testified that the wavelet is a two-dimensional dyadic wavelet. The computation of the wavelet transform in image processing depends only upon the wavelet functional values at a few integer points. This can make quite computing true. Through many experiments on detection of fabric composite's snapped cross sections, clearer and more consistent results are obtained than traditional methods. Computing fiber volume fraction of fabric composites on it, we can get the satisfactory results, which has magnificent significance on the study of the properties of the fabric composite. Meanwhile, those results are in accordance with the data supplied by manufacture. This contributes greatly to the automatic detection of the fabric composite's parameters.