CAJ | 학술논문

针对汽车轮胎生产线检测的智能化管理问题,提出一种嵌入式系统以对轮胎规格号进行快速识别.对图像进行预处理,包括极坐标变换、双线性插值拉伸,利用大津法进行二值化,以连通域方法来提取规格号字符.采用基于统计的方法抽取轮廓特征进行规格号模式识别.系统硬件部分采用高速TMS320C6000系列DSP为核心的嵌入式硬件系统,使用参数预存储、软件流水线技术及双缓冲内存分配等优化方法.实验结果表明,该系统能快速有效地识别汽车轮胎规格号.
침대기차륜태생산선검측적지능화관리문제,제출일충감입식계통이대륜태규격호진행쾌속식별.대도상진행예처리,포괄겁좌표변환、쌍선성삽치랍신,이용대진법진행이치화,이련통역방법래제취규격호자부.채용기우통계적방법추취륜곽특정진행규격호모식식별.계통경건부분채용고속TMS320C6000계렬DSP위핵심적감입식경건계통,사용삼수예존저、연건류수선기술급쌍완충내존분배등우화방법.실험결과표명,해계통능쾌속유효지식별기차륜태규격호.
For Intelligent management of car tire production line, a method of fast recognition of car tire size code which combining a drop-in handware system is presented. Tire image is preprocessed including polar transformation, bilinear interpolation stretching and Da-Jing Method to preprocess tire image, segments specifications by connective region. Outline features of size code is extracted based on statistical methods to recognize tire size code. In the drop-in handware system, TMS320C6000 DSP is applied to realize this algorithm, program can be executed quickly by using methods such as soft assembly line, parameter storage and two-buffer memory allocation technology. Experimental result shows that the system method meets the needs of fast recognition of car tire size code.