基于FPGA的視頻自適應(yīng)Sobel算法設(shè)計(jì)與實(shí)現(xiàn)

摘  要：該文首先對(duì)標(biāo)準(zhǔn)Sobel算法進(jìn)行分析，然后針對(duì)算法在其他方向邊緣檢測(cè)效果不好和閾值需要人為設(shè)定等不足對(duì)算法進(jìn)行改進(jìn)和優(yōu)化。該設(shè)計(jì)采用FPGA作為硬件平臺(tái)，利用其并行處理能力，對(duì)Sobel算法進(jìn)行硬件加速，使其能夠自適應(yīng)地處理不同分辨率和幀率的視頻流。

關(guān)鍵詞：Sobel算法；雙線性插值；FPGA；自適閾值；硬件加速

中圖分類號(hào)：TP301      文獻(xiàn)標(biāo)志碼：A          文章編號(hào)：2095-2945（2024）32-0015-04

Abstract： This paper first analyzes the standard Sobel algorithm， and then improves and optimizes the algorithm to address the shortcomings of the algorithm such as poor edge detection performance in other directions and the need to manually set thresholds. This design uses FPGA as the hardware platform， and uses its parallel processing capabilities to hardware accelerate Sobel algorithm， making it able to adaptively process video streams with different resolutions and frame rates.

Keywords： Sobel algorithm; bilinear interpolation; FPGA; adaptive threshold; hardware acceleration

邊緣檢測(cè)是找出圖像中物體的邊界，這些邊界通常表現(xiàn)為像素強(qiáng)度的顯著變化[1-2]。（剩余3047字）