Unlock instant, AI-driven research and patent intelligence for your innovation.

A face detection method based on arm Cotex-A series platform

A face detection and implementation method technology, applied in the direction of processor architecture/configuration, program code conversion, instruments, etc., can solve the problems of low detection efficiency and unusable acceleration methods, and achieve the effect of improving efficiency

Active Publication Date: 2021-12-24
SHISHI SENKE INTELLIGENT TECH CO LTD
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The main problem of using seetaface now is: when using seetaface for face detection, the x86 platform can use SSE instructions to accelerate, but under the limitation of the instruction set under the ARM platform, the original acceleration method cannot be used, so the detection efficiency is low

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A face detection method based on arm Cotex-A series platform
  • A face detection method based on arm Cotex-A series platform
  • A face detection method based on arm Cotex-A series platform

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0058] Step S1, modify the source code in FaceDetection of seetaface under the hardware environment of ARM Cotex-A series processors, and modify the compiler type to be a cross compiler:

[0059] Modify the SET command,

[0060] SET(CMAKE_SYSTEM_NAME Linux)————①

[0061] SET(CMAKE_CXX_COMPILER

[0062] " / opt / hisiv-linux / x86-arm / arm-hisiv400-linux / bin / arm-hisiv400-linux-gnueabi-g++")————②

[0063] ①: Set the system type. Choose to use linux, which must be set when using a cross compiler;

[0064] ②: Set the cross compiler path. Enable the cross compiler and add the path of the cross compiler;

[0065] Step S2, add NEON compilation options in the compiler settings:

[0066] Modify the relevant settings of enabling instructions in cmakelist.txt, enable NEON, and modify (increase) the compilation options of NEON in the compiler option settings in the set command.

[0067] -mfloat-abi=softfp-mfpu=neon.

[0068] Step S3, replace the header file required by the original SSE i...

Embodiment 2

[0102] Such as image 3 As shown, when using the method of the present invention, first process the feature points of the input image to determine whether vector calculation is required, and if necessary, modify the code according to the method of Embodiment 1.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a method for implementing face detection based on ARM Cotex-A series platforms, comprising the following steps: S1, modifying the source code in the FaceDetection of seetaface under the hardware environment of ARM Cotex-A series processors, and modifying the compiler type to Cross-compiler; S2, add NEON compilation options in the compiler settings; S3, replace the header files required by the original SSE instructions in FaceDetection with the header files required by NEON; S4, use SSE instructions in the original code for FaceDetection Partially modify it into NEON instructions, and modify the functions using SSE instructions into functions using NEON; S5, under the support of the compilation option of NEON, which has been added in step S2, recompile the program to obtain the required dynamic link library file, thereby compiling to obtain the The FaceDetection program that supports NEON under the ARM Cotex‑A series processor platform. The invention can effectively improve the efficiency of face detection using seetaface in hardware using Cotex-A series processors.

Description

technical field [0001] The invention relates to the field of computer image processing, in particular to a method for realizing face detection on an ARM-based Cotex-A series platform. Background technique [0002] The existing face detection has certain requirements on the image processing equipment (hardware), so the face detection module of seetaface is mostly used in the x86 platform. The detection process is as follows: figure 1 shown. ARM processors are mostly used in mobile devices and embedded devices. This kind of hardware adopts some balance between energy consumption and performance, so the performance is not very high, but limited by performance, face detection in these hardware environments The efficiency is not high. [0003] SIMD Single Instruction Multiple Data (Single Instruction Multiple Data, SIMD) is a method that uses one controller to control multiple processors while performing the same operation on each of a set of data (also known as "data vector") ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06T1/20G06F8/41
CPCG06F8/41G06T1/20G06V40/16
Inventor 洪朝群王善炮
Owner SHISHI SENKE INTELLIGENT TECH CO LTD