条码技术目前已经深入到生活的各个领域,条码扫描系统产生的扫描信号的处理仍然是研究的焦点。目前有很多种方法对条码图像进行去噪和识别处理,但去除噪声的效果不尽理想,尤其是对高斯白噪声的去除值是在噪声估计的基础上进行,由于噪声的随机行决定了对噪声的实时估计存在着误差,影响了实际的去噪效果。这里并没有对噪声信号进行估计,而是将扫描系统的扫描采样信号进行双谱分析,在双谱分析下用数学方式理论上去除高斯白噪声从而得到纯净的扫面信号,通过仿真实验证明,在SNR较低的状态下对采用该方法去除高斯白噪声有非常理想的处理效果。
The technology of bar code has been used in different domains. The method of dealing with bar code signal is the focus of studying. In recent years there are many methods on denoising, but they can not get ideal result. Especially for removing Ganssian white noise, the noise is estimated. Because the noise is random, the estimated noise is not real time noise. The actual effects are affected. The method in this paper is used to denoise by bispectrum analysis. It can get real signal that we want. The sampled siznals are analysed by bispectrum. In lower SNR condition, the method can get better real signal.