城市建筑色彩属性信息可视化挖掘系统设计

Design of visual mining system for urban building color attribute information

传统的建筑色彩属性信息可视化挖掘系统无法降低建筑色彩属性信息中包含的数据量,导致占用了较高的系统内存和信道,降低了属性信息挖掘程度,因此该文设计一种城市建筑色彩属性信息可视化挖掘系统。系统的硬件设计中,设计了多通道边缘融合控制器及其工作流程,优化了多角度图片的拼接,为色彩属性信息挖掘提供基础;软件设计中,在RGB色彩空间中对图片进行标准化预处理,实现图片在不同色彩空间的转换,精简属性信息中的数据量,通过小波变换图像分解,DCT变换提取出色彩的压缩编码,以减轻通信信道压力。在系统测试的实验中,分别使用该文设计系统和原有系统进行峰值信噪比测试,结果表明,该文设计的系统具有更高的峰值信噪比,能够提高色彩属性信息可视化的挖掘程度。

As the traditional visual mining system of building color attribute information cannot reduce the amount of data contained in the building color attribute information,which results in the high occupation of system memory and channel,and reduces the degree of attribute information mining,a visual mining system of urban building color attribute information is designed.In the hardware design of the system,the multi⁃channel edge fusion controller and its working process are designed,and the splicing of the multi⁃angle images is optimized,which provide the foundation for the color attribute information mining.In the software design,the standardized preprocessing of images is carried out in RGB color space to realize the image conversion in different color spaces.In order to simplify the amount of data in the attribute information,the wavelet transform is required for image decomposition,and the DCT transform is used to extract the color compressed encoding,so as to reduce the burden of communication channel.In the experiment of system test,the designed system and the original system were respectively used to test the PSNR(peak signal to noise ratio).The results show that the designed system has a higher PSNR and can improve the mining degree of color attribute information visualization.