CAJ | 학술논문

针对非均匀光照条件下图像的分割问题，提出一种基于FPGA的实时图像分割方法。得到了一种非均匀光照下分割效果较好的方法。首先，使用图像传感器采集原始的图像数据，然后进行so-bel边缘检测，最后对其进行Bernsen局部二值化得到分割图像。本方法使用FPGA来对图像进行高速采集和并行处理，实时完成图像边缘检测及局部二值化。经现场调试证明本方法可以有效分割图像，使其目标更加突出，从而改善了非均匀光照对图像分割质量带来的影响，并具有良好的实时性。
침대비균균광조조건하도상적분할문제，제출일충기우FPGA적실시도상분할방법。득도료일충비균균광조하분할효과교호적방법。수선，사용도상전감기채집원시적도상수거，연후진행so-bel변연검측，최후대기진행Bernsen국부이치화득도분할도상。본방법사용FPGA래대도상진행고속채집화병행처리，실시완성도상변연검측급국부이치화。경현장조시증명본방법가이유효분할도상，사기목표경가돌출，종이개선료비균균광조대도상분할질량대래적영향，병구유량호적실시성。
Aiming at the problem of image segmentation in the condition of non-uniform illumination,a real-time image capture and segmentation method based on FPGA is proposed,which achieved a much better result in the condition. Firstly,the image sensor captures original image data. Secondly,the edge of the image is detected by so-bel algorithm and the outstanding subject image is achieved. Thirdly,the Bernsen local binarization is applied to process the subject image,and then the segmented image is obtained. This method adopts FPGA to capture image <br> quickly and process image data parallel. Besides,image edge detection and local binarization are completed in real time. The result shows that the method can make the subject image more outstanding and improve the quality of im-age segmentation greatly under the condition of non-uniform illumination. In addition,the real-timing of the method is good.