CAJ | 학술논문

提出了一种基于ARM嵌入式系统的自动智能视频监控报警系统设计方案，改进了一种基于图像RGB到HLS色彩空间变换的图像处理算法，有效地解决了图像自动识别过程中亮度对于识别精度的影响，并对复杂的HLS色彩空间变换算法进行了优化，使之能够流畅运行在当前流行的ARM嵌入式系统上。提出了一种摄像头位移自动判断识别方法，能够有效地解决摄像头角度和位置变化导致的图像基准背景变化对自动识别精度带来的巨大影响。该视频图像处理技术能够应用于智能小区住户以及中小型企业的安全保障系统，并能显著降低安保投入成本，提升安保准确率。
제출료일충기우ARM감입식계통적자동지능시빈감공보경계통설계방안，개진료일충기우도상RGB도HLS색채공간변환적도상처리산법，유효지해결료도상자동식별과정중량도대우식별정도적영향，병대복잡적HLS색채공간변환산법진행료우화，사지능구류창운행재당전류행적ARM감입식계통상。제출료일충섭상두위이자동판단식별방법，능구유효지해결섭상두각도화위치변화도치적도상기준배경변화대자동식별정도대래적거대영향。해시빈도상처리기술능구응용우지능소구주호이급중소형기업적안전보장계통，병능현저강저안보투입성본，제승안보준학솔。
This paper presents an algorithm and implementation about automatic video monitoring ,and an image processing algorithm based on the images RGB to HSV color space conversion is proposed . It can remove the impact of the brightness on the image recognition ,and optimize the HSV color space conversion algorithm in order to run in the current popular ARM Embedded System smoothly . At last , this paper proposes a method to eliminate the adverse impact which is generated by background change because of moving camera .This image processing algorithm can be applied to the security system of intelligence community ’s houses and small and medium enterprise company to reduce investment and increase accuracy significantly .