噪声在理论上可以定义为“不可预测，只能用概率统计方法来认识的随机误差”。实际获得的图像一般都因受到某种干扰而含有噪声。引起噪声的原因有敏感元器件的内部噪声、相片底片上感光材料的颗粒、传输通道的干扰及量化噪声等。(In theory, noise can be defined as "unpredictable, random errors that can only be recognized by probability and statistics". The actual images are usually caused by some kind of interference and contain noise. The noise is caused by the internal noise of the sensitive components, the particles of the photoreceptor on the film, the interference of the transmission channel and the quantization noise.)