• INTRODUCTION
• INSTALLATION
• METHODS and Examples

INTRODUCTION

This package is to identify person faces by giving training images and test images. It is simple library. This library require installation of opencv in your system.opencv4nodejs

INSTALLATION

To install opencv4nodejs you need following dependency/installation to compile c++ library. FOR MORE DETAILS VISIT opencv4nodejs

Install this package by npm i face-identity --save

1.cmake 2.git 3.windows build tools

In last step perform npm rebuild

METHODS/EXAMPLES

Recognizer values (last parameter of function) *pass in lowercase*
1.lbph
2.eigen
3.fisher


1.FaceDetectFS(DIR, MatchImageBase64, SelectedRecognizer)

Explanation: 
This method takes 3 parameters 
<b>DIR:</b> You need to specify directory containing training image folders.
<b>MatchImageBase64:</b> Pass base64 data of a test image.
<b>SelectedRecognizer:</b> Recognizer values (lbph,eigen and fisher)


*For this function you need to store images of persons in separate folders *

i.e. Consider you have 5 persons faces Eve,Adam,James,Mike and Tyson and you have base64 image data of Adam

so your file/directory structure should be 

Root directory E:\CODE\Face-Identity\

Training Data
<b>E:\CODE\Face-Identity\Eve\ <All image files of Eve>
E:\CODE\Face-Identity\Adan\ <All image files of Adam>
E:\CODE\Face-Identity\James\ <All image files of James>
E:\CODE\Face-Identity\Mike\ <All image files of Mike>
E:\CODE\Face-Identity\Tyson\ <All image files of Tyson></b>

So it will take image data from all directories and return prediction <Directory Name and probability> 

const faceid = require('face-identity');
const fs = require('fs')
const path = require('path');

const base=fs.readFile(path.join(__dirname,"/base64.txt"),'utf8',(err,data)=> //Reading textfile containing base64 of an image
{
    const predictionFs =faceIdentity.FaceDetectFS(path.join(__dirname,"/Face-Identity"),data,"lbph");
    console.log(predictionFs);
});



2.FaceDetectFSSearch(SearchDIR, PersonNames, MatchImageBase64, SelectedRecognizer)

Explanation: 
This method takes 4 parameters 
<b>SearchDIR:</b> You need to specify directory containing training image folders.
<b>PersonNames:</b> Specify Person names (Pass Array).
<b>MatchImageBase64:</b> Pass base64 data of a test image.
<b>SelectedRecognizer:</b> Recognizer values (lbph,eigen and fisher)

Prerequisite : ALL THE IMAGES MUST HAVE PERSON NAME AS A FILE NAME
This method search images based on Person Name in given directory and train the model.  

Example 

Root directory E:\CODE\Face-Identity\<All the images of persons(Eve,Adam,James,Mike and Tyson)>



const faceid = require('face-identity');
const fs = require('fs')
const path = require('path');
const base=fs.readFile(path.join(__dirname,"/base64.txt"),'utf8',(err,data)=>
{
    const predictionFsSearch =faceIdentity.FaceDetectFSSearch(path.join(__dirname,"/face-recognition"),['Eve','Adam','James','Mike' ,'Tyson'],data,"lbph");
    console.log(predictionFsSearch);

});