CAJ | 학술논문

为了弥补市场上现有印鉴鉴别系统体积大、移动性能差、安全性较低、价格比较昂贵等缺陷,研究了基于高速数字信号处理器(DSP)和现场可编程门阵列(FPGA)的嵌入式印鉴鉴别系统.系统在印鉴识别算法上,基于平滑卷积的方法计算印鉴的中心位置和半径.采用径向投影法,对一维特征数据计算,得到待验印鉴(SS)与预留印鉴(MS)之间的偏转角度.把印鉴质量指标作为MS与SS对应边缘Hausdorff距离测度的控制参数,用神经网络方法综合分析、判别印鉴真伪.在硬件实现方面,片上可编程(SOPC)系统结合DSP作为检测系统核心.SOPC系统包括控制器、图像预处理器等.DSP作为系统的主处理器,用于进行图像的特征检测与识别.系统具有以太网、RS232、USB等通用接口.实验表明,该系统可以有效识别印鉴,并具有体积小、成本低、系统功能可灵活升级等特点.
위료미보시장상현유인감감별계통체적대、이동성능차、안전성교저、개격비교앙귀등결함,연구료기우고속수자신호처리기(DSP)화현장가편정문진렬(FPGA)적감입식인감감별계통.계통재인감식별산법상,기우평활권적적방법계산인감적중심위치화반경.채용경향투영법,대일유특정수거계산,득도대험인감(SS)여예류인감(MS)지간적편전각도.파인감질량지표작위MS여SS대응변연Hausdorff거리측도적공제삼수,용신경망락방법종합분석、판별인감진위.재경건실현방면,편상가편정(SOPC)계통결합DSP작위검측계통핵심.SOPC계통포괄공제기、도상예처리기등.DSP작위계통적주처리기,용우진행도상적특정검측여식별.계통구유이태망、RS232、USB등통용접구.실험표명,해계통가이유효식별인감,병구유체적소、성본저、계통공능가령활승급등특점.
An embedded seal imprint identifiction system based on high speed digital signal processor (DSP) and field-programmable gate array (FPGA) is designed and fabricated. In recognition algorithms of the system, spatial coordinates of centers and radius of tested seals are calculated with a method based on flatness convolution. The orientation between sample seal (SS) and model seal (MS) is calculated with the radius projection method, which calculates a group of one-dimensiond feature data instead of two-dimensiond. Improved Hausdorff distance is used to measure the similarity between seals, as seal quality being the control parameters. Artificial neural network is adopted in analyzing and estimation. In respect of hardware realization, System-on-a-programmable chip (SOPC)combining DSP serves as CPU of this verification system. Image preprocessor and controller are included in the SOPC system. As the master processor, DSP is applied to accomplish image feature detection and verification. General purpose interface (GPI) such as Ethernet interface, USB and RS232 is designed in the system. Experiment shows that seals can be effectively identified and the function of the small-size and low-cost verification system can be upgraded flexibly.