通用学习式红外遥控器中数据压缩与识别算法

Data compression and pattern recognition algorithm in a universal infrared remote control transmitter

对数据进行有效的压缩与识别是系统的关键所在。针对红外遥控信号的数据特点 ,提出了一种数据压缩与识别算法。该算法将遥控码流归纳为 3种模式 ,用矢量量化编码方法进行数据压缩 ,采用聚类方法解决了由时钟精度带来的误差问题 ,并通过识别发送码流的模式进一步压缩存储空间。该算法大大提高了识别与存储效率 ,并已成功应用于两种遥控器产品中。

The universal infrared remote control transmitter is a new type of remote control transmitter which can learn remote signals. The critical problem of the transmitter system is how to compress the data and to recognize the signal pattern. This paper presents a data compression and pattern recognition algorithm for the universal remote control transmitter. The remote control code flow is reduced to three patterns. Vector quantization coding and the cluster algorithm are used to compress the data memory space more by recognizing code flow patterns. This algorithm, which increases the recognition reliability and the memory usage efficiency, has been successfully used in two kinds of universal infrared remote control transmitters.