Sobel Filter - 2023.1 English

The Sobel function Computes the gradients of input image in both x and y direction by convolving the kernel with input image being processed.

For Kernel size 3x3
- GradientX:
- GradientY:
For Kernel size 5x5
- GradientX:
- GradientY:
For Kernel size 7x7
- GradientX:
- GradientY:

API Syntax

template<int BORDER_TYPE,int FILTER_TYPE, int SRC_T,int DST_T, int ROWS, int COLS,int NPC=1, bool USE_URAM=false, int XFCVDEPTH_IN = _XFCVDEPTH_DEFAULT, int XFCVDEPTH_OUT_X = _XFCVDEPTH_DEFAULT, int XFCVDEPTH_OUT_Y = _XFCVDEPTH_DEFAULT>
void Sobel(xf::cv::Mat<SRC_T, ROWS, COLS, NPC, XFCVDEPTH_IN> & _src_mat,xf::cv::Mat<DST_T, ROWS, COLS, NPC, XFCVDEPTH_OUT_X> & _dst_matx,xf::cv::Mat<DST_T, ROWS, COLS, NPC, XFCVDEPTH_OUT_Y> & _dst_maty)

Parameter Descriptions

The following table describes the template and the function parameters.

Table 705 Table . Sobel Parameter Description
Parameter	Description
FILTER_TYPE	Filter size. Filter size of 3 (XF_FILTER_3X3), 5 (XF_FILTER_5X5) and 7 (XF_FILTER_7X7) are supported.
BORDER_TYPE	Border Type supported is XF_BORDER_CONSTANT
SRC_T	Input pixel type. Only 8-bit, unsigned, 1 and 3 channels are supported (XF_8UC1 and XF_8UC3)
DST_T	Output pixel type. Only 8-bit unsigned, 16-bit signed,1 and 3 channels are supported (XF_8UC1, XF_16SC1,XF_8UC3 and XF_16SC3)
ROWS	Maximum height of input and output image.
COLS	Maximum width of input and output image. Must be multiple of 8, for 8-pixel operation.
NPC	Number of pixels to be processed per cycle; possible options are XF_NPPC1 and XF_NPPC8 for 1 pixel and 8 pixel operations respectively.
USE_URAM	Enable to map storage structures to UltraRAM
XFCVDEPTH_IN	Depth of the input image.
XFCVDEPTH_OUT_X	Depth of the output image.
XFCVDEPTH_OUT_Y	Depth of the output image.
_src_mat	Input image
_dst_matx	X gradient output image.
_dst_maty	Y gradient output image.
Sobel 7x7 8-pixel is not supported.

Resource Utilization

The following table summarizes the resource utilization of the kernel in different configurations, generated using Vivado HLS 2019.1 tool for the Xilinx Xczu9eg-ffvb1156-1-i-es1 FPGA, to process a grayscale HD (1080x1920) image.

Table 706 Table . Sobel Function Resource Utilization Summary
Operating Mode	Filter Size	Operating Frequency (MHz)	Utilization Estimate
Operating Mode	Filter Size	Operating Frequency (MHz)	BRAM_18K	DSP_48Es	FF	LUT	CLB
1 pixel	3x3	300	3	0	609	616	135
	5x5	300	5	0	1133	1499	308
	7x7	300	7	0	2658	3334	632
8 pixel	3x3	150	6	0	1159	1892	341
	5x5	150	10	0	3024	5801	999

Table 707 Table . Sobel Function Resource Utilization Summary
Operating Mode	Filter Size	Operating Frequency (MHz)	Utilization Estimate
Operating Mode	Filter Size	Operating Frequency (MHz)	BRAM_18K	DSP_48Es	FF	LUT
1 pixel	3x3	300	18	0	1047	1107
	5x5	300	30	0	5370	3312
	7x7	300	42	0	6100	5496

The following table summarizes the resource utilization of the kernel in different configurations, generated using Vivado HLS 2019.1 tool for the Xilinx xczu7ev-ffvc1156-2-e FPGA, to process a grayscale 4K (3840x2160) image with UltraRAM enable.

Table 708 Table . Sobel Function Resource Utilization Summary with UltraRAM enable
Operating Mode	Filter Size	Operating Frequency (MHz)	Utilization Estimate
Operating Mode	Filter Size	Operating Frequency (MHz)	BRAM_18K	URAM	DSP_48Es	FF	LUT
1 pixel	3x3	300	0	1	0	919	707
	5x5	300	0	1	0	2440	1557
	7x7	300	0	1	0	4066	3495
8 pixel	3x3	150	0	3	0	1803	2050
	5x5	150	0	5	0	4159	6817

Performance Estimate

The following table summarizes the performance of the kernel in different configurations, as generated using Vivado HLS 2019.1 tool for the Xilinx Xczu9eg-ffvb1156-1-i-es1, to process a grayscale HD (1080x1920) image.

Table 709 Table . Sobel Function Performance Estimate Summary
Operating Mode	Operating Frequency (MHz)	Filter Size	Latency Estimate (ms)
1 pixel	300	3x3	7.5
	300	5x5	7.5
	300	7x7	7.5
8 pixel	150	3x3	1.7
8 pixel	150	5x5	1.71