Wavelet transform algorithm of seam image process with DSP

The position of seam center was got with template match algorithm. First seam template was acquired through wavelet transform, then the template was shifted in subsequent seam image and correlation coefficients were calculated accordingly, the column corresponding to the max correlation coefficient was taken as seam center position of subsequent seam image. The seam image is processed with TDS642EVM since DSP has advantages in digital signal processing. Principle of 2-D wavelet transform algorithm was analyzed, then wavelet transform programs on DSP are developed. At last, example of processing pipeline seam image with TDS642EVM was given.