Page 1 of 21

futomi / node-omron-hvc-p2

Dismiss
Join GitHub today
GitHub is home to over 28 million developers working together to host
and review code, manage projects, and build software together.

Sign up

The node-omron-hvc-p2 is a Node.js module which allows you to communicate with the image sensing device "OMRON
Human Vision Components (HVC-P2)" through a USB serial port.

7 commits 1 branch 0 releases 1 contributor MIT

Branch: master New pull request Find file Clone or download

futomi Released v0.1.2 Latest commit 61ec216 on Nov 4 2017

images First Public Release a year ago

lib Released v0.1.2 10 months ago

.gitattributes Added .gitattributes & .gitignore files a year ago

.gitignore Released v0.1.1 11 months ago

.npmignore Released v0.1.1 11 months ago

LICENSE First Public Release a year ago

README.md Released v0.1.2 10 months ago

README_ja.md Released v0.1.2 10 months ago

package.json Released v0.1.2 10 months ago

README.md

node-omron-hvc-p2
[Japanese (日本語)]

The node-omron-hvc-p2 is a Node.js module which allows you to communicate with the image sensing device "OMRON
Human Vision Components (HVC-P2)" through a USB serial port.

The HVC-P2 supports:

• Human Body Detection

• Hand Detection
• Face Detection
• Face Direction Estimation
 Page 2 of 22

• Age Estimation
• Gender Estimation
• Gaze Estimation
• Blink Estimation
• Expression Estimation
• Face Recognition

The node-omron-hvc-p2 supports all functions supported by the HVC-P2. Besides, it allows you to save a camera view image
as an image file such as GIF, JPEG, or PNG.

Dependencies
• Node.js 6 +
• serialport 5.0.0 +
◦ If the serialport module has been already installed in you host computer, check the version. The node-omron-hvc-p2
now does not support older versions of the serialport module than 5.0.0 .
• node-gd (Optional, for Linux, Mac)
• lwip (Optional, for Windows)

Installation

$ cd ~
$ npm install serialport
$ npm install node-omron-hvc-p2

The node-omron-hvc-p2 requires an image processing module in order to create images captured by the HVC-P2. For Linux
and Mac, the node-gd is required. For Windows, the lwip is required.

If you don't need to get images captured by the HVC-P2, the image processing modules are not required.

Debian/Ubuntu
It is recommended to install node-gd.

$ sudo apt-get install libgd2-dev # libgd

$ npm install node-gd

RHEL/CentOS
It is recommended to install node-gd.

$ sudo yum install gd-devel

$ npm install node-gd

Mac OS
It is recommended to install node-gd.

$ sudo port install pkgconfig gd2

$ npm install node-gd

Windows
It is recommended to install lwip.

$ npm install lwip

Table of Contents
 Page 3 of 21

• Quick Start
◦ Detecting faces
◦ Getting an image
• HvcP2 object
◦ connect([params]) method
◦ disconnect() method
◦ getSerialPortPath() method
◦ getModelVersion([params]) method
◦ detect(params) method
Face detection
Body detection
Hand detection
◦ getConfigurations([params]) method
◦ setConfigurations(configrations) method
◦ resetConfigurations() method
◦ getFaceRecognitionData(params) method
◦ getFaceRecognitionUsers([params]) method
◦ addFaceRecognitionData(params) method
◦ deleteFaceRecognitionData(params) method
◦ deleteFaceRecognitionUser(params) method
◦ clearFaceRecognitionData() method
◦ saveAlbum(params) method
◦ loadAlbum(params) method
◦ saveAlbumOnFlashROM() method
◦ reformatFlashROM() method
• Release Note
• References
• License

Quick Start

Detecting faces
This sample code shows how to connect to the HVC-P2 and how to send a command to detect faces.

// Load the node-omron-hvc-p2 and get a `HvcP2` constructor object

const HvcP2 = require('node-omron-hvc-p2');
// Create a `HvcP2` object
const hvcp2 = new HvcP2();

// Connect to the HVC-P2

hvcp2.connect().then(() => {
// Send a command for detecting
return hvcp2.detect({
face : 1, // Enable face detection
age : 1, // Enable age estimation
gender : 1 // Enable gender Estimation
});
}).then((res) => {
// Show the result
console.log(JSON.stringify(res, null, ' '));
// Disconnect the HVC-P2
return hvcp2.disconnect();
}).then(() => {
console.log('Disconnected.');
}).catch((error) => {
console.error(error);
});

First of all, you have to create a HvcP2 object from the HvcP2 constructor object. In the code above, the variable hvcp2 is the
HvcP2 object.
 Page 4 of 21

Calling the connect() method, the node-omron-hvc-p2 starts to find a USB serial port which the HVC-P2 is connected, then
prepares for use.

Calling the detect() method, the node-omron-hvc-p2 sends a command to the HVC-P2 in order to detect faces in the camera
view. The HVC-P2 supports a lot of detect options. In the code above, Face detection, age estimation, gender estimation are
enabled.

The sample code above will output the result as follows:

{
"face": [
{
"face": {
"x": 930,
"y": 497,
"size": 390,
"confidence": 579
},
"age": {
"age": 41,
"confidence": 500
},
"gender": {
"gender": 1,
"confidence": 1000
}
}
]
}
Disconnected.

The result says that a face was detected, the age is 41 years old, and the gender is male.

Finally, you can disconnect the device using disconnect() method.

Getting an image
The detect() method can return a image data as well. In the code blow, some parameters related the image option are
passed to the method.

// Load the node-omron-hvc-p2 and get a `HvcP2` constructor object

const HvcP2 = require('node-omron-hvc-p2');
// Create a `HvcP2` object
const hvcp2 = new HvcP2();

// Connect to the HVC-P2

hvcp2.connect().then(() => {
// Send a command for detecting
return hvcp2.detect({
face : 1, // Enable face detection
age : 1, // Enable age estimation
image : 1, // Enable capturing image
imageType : 3, // Save the image as a file
imageFormat : 'png', // Image format
imagePath : './test.png', // File path
imageMarker : true // Draw markers in the image
});
}).then((res) => {
// Show the result
console.log(JSON.stringify(res, null, ' '));
// Disconnect the HVC-P2
return hvcp2.disconnect();
}).then(() => {
console.log('Disconnected.');
}).catch((error) => {
console.error(error);
});

The sample code above will output the result as follows:

{
"face": [
 Page 5 of 21

{
"face": {
"x": 624,
"y": 588,
"size": 330,
"confidence": 559
},
"age": {
"age": 25,
"confidence": 666
}
}
],
"image": {
"width": 320,
"height": 240
}
}
Disconnected.

Besides, an image file will be created in the current directory. If the script is run on Linux or Mac, the image will be like this:

If the detected face was recognized as a woman, the color of markers is red. If recognized as a man, the color is blue.

If the script is run on Windows, the image will be like this:

Note that, as you can see, the age is not supported for Windows.

HvcP2 object

In order to use the node-omron-hvc-p2, you have to load the node-omron-hvc-p2 module as follows:

const HvcP2 = require('node-omron-hvc-p2');

You can get a HvcP2 constructor from the code above. Then you have to create an HvcP2 object from the HvcP2 constructor
as follows:

const hvcp2 = new HvcP2();

In the code snippet above, the variable hvcp2 is a HvcP2 object. The HvcP2 object has methods as described in sections
below.
 Page 6 of 21

connect([params]) method
The connect() method finds the HVC-P2 connected to a USB port on your host PC, then prepares for use. This method returns
a Promise object.

hvcp2.connect().then(() => {
console.log('Connected.');
// Do something.
}).catch((error) => {
console.error(error);
});

Basically, you don't need to know which USB serial port your HVC-P2 is connected. This method try to find an appropriate USB
serial port. Besides, you don't need to specify the baud rate. This probably works well on at least Debian-based Linux distro
(e.g., Raspbian, Ubuntu), Mac, and Windows.

However, the node-omron-hvc-p2 does not necessarily find an appropriate USB serial port. You can specify the USB serial port
and the baud rate.

hvcp2.connect({
path: 'COM4',
baudRate: 921600
}).then(() => {
console.log('Connected.');
// Do something.
}).catch((error) => {
console.error(error);
});

The connect() method takes a hash object as an argument containing properties as follows:

Property Required Type Description

The path representing the serial port which your HVC-P2 is connected. (e.g.,
path optional String
"COM3", "/dev/ttyACM0", "/dev/tty-usbserial1")

baudRate optional Number 9600 , 38400 , 115200 , 230400 , 460800 , or 921600 (Default)

If you want to finish the connection process as soon as possible, you should specify the property path because the auto scan
mode (no path specified) takes a little more time depending on the environment of your host PC.

disconnect() method
The disconnect() method releases the connection with your HVC-P2 on the USB port. This method returns a Promise object.

hvcp2.disconnect().then(() => {
console.log('Disconnected.');
}).catch((error) => {
console.error(error);
});

getSerialPortPath() method
The getSerialPortPath() method returns the path representing the USB serial port assigned to your HVC-P2. Note that this
method does not return a Promise object unlike other methods.

hvcp2.connect().then(() => {
console.log('Serial Port Path: ' + hvcp2.getSerialPortPath());
}).catch((error) => {
console.error(error);
});

The code above will output the result as follows if the code is run on Windows:

Serial Port Path: COM4

 Page 7 of 21

If the code is run on Raspbian, the result will be:

Serial Port Path: /dev/ttyACM0

Note that this method returns an empty string if any HVC-P2 is not connected using the connect() method.

getModelVersion([params]) method
The getModelVersion() method reports the model name and the version number of your HVC-P2. This method returns a
Promise object. This method takes a hash object as an argument containing properties as follows:

Property Type Required Description

true (default) or false . If true is specified or this property is not specified, this
cache Boolean Optional method reports the data cached when the connect() method was called. If false is
specified, this method requests to the HVC-P2.

The node-omron-hvc-p2 cashes the model name and version number when the connect() method is called. This method
reports the cashed information by default because the model name and the version number never change. But if you want to
make a query again, specify the property cache and set it to false .

The code blow calls this method and show the cached information.

hvcp2.getModelVersion().then(() => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The code below calls this method and show the newly queried information.

hvcp2.getModelVersion({cache: false}).then(() => {

console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The codes above will output the result as follows:

{
"model": "B5T-007001",
"major": 1,
"minor": 0,
"release": 1,
"revision": 9372
}

detect(params) method
The detect() method makes your HVC-P2 detect objects in the camera view. This method returns a Promise object. This
method takes a hash object as an argument containing properties as follows:

Property Type Required Description

body Number optional Human Body Detection ( 0 : disable (default), 1 : enable)

hand Number optional Hand Detection ( 0 : disable (default), 1 : enable)

face Number optional Face Detection ( 0 : disable (default), 1 : enable)

direction Number optional Face Direction Estimation ( 0 : disable (default), 1 : enable)

age Number optional Age Estimation ( 0 : disable (default), 1 : enable)

gender Number optional Gender Estimation ( 0 : disable (default), 1 : enable)

gaze Number optional Gaze Estimation ( 0 : disable (default), 1 : enable)

 Page 8 of 21

Property Type Required Description

blink Number optional Blink Estimation ( 0 : disable (default), 1 : enable)

expression Number optional Expression Estimation ( 0 : disable (default), 1 : enable)

recognition Number optional Face Recognition ( 0 : disable (default), 1 : enable)

image Number optional Image output ( 0 : disable (default), 1 : 320x240 pixel, 2: 160x120 pixel)

imageType Number optional 0 : Array (default), 1 : Buffer, 2 : Data URL, 3 : File

imageFormat String optional "gif" (default), "jpg" , or "png"

imagePath String conditional File path with file name (e.g., "/tmp/image.png" )

imageMarker Boolean optional true or false (default)

Though all properties are optional, at least one among body , hand , face , direction , age , gender , gaze , blink ,
expression , recogunition , and image must be set to 1 .

The imagePath is required if the image is 1 or 2 and the imageType is 3 (File).

The result of the detection will be pass to the resolve() function, which is a hash object containing the properties as follows:

Property Type Description

The result of body detection. This property exists only when the request parameter body is 1 .
body Array
Otherwise, this property does not exists. See the section "Body detection" for details.

The result of hand detection. This property exists only when the request parameter hand is 1 .
hand Array
Otherwise, this property does not exists. See the section "Hand detection" for details.

The result of face detection. This property exists only when at least one of request parameters
face Array among face , direction , age , gender , gaze , blink , expression , and recognition is 1 .
Otherwise, this property does not exists. See the section "Face detection" for details.

The image data. This property exists only when the request parameter image is 1 or 2 .
image Object
Otherwise, this property does not exists.

+ width Number The width of the image (pixel).

+ height Number The height of the image (pixel).

+ pixels Array The pixcel data of the image. This property exists only if the request parameter imageType is 0 .

The Buffer object representing the image. This property exists only if the request parameter
+ buffer Buffer
imageType is 1 .

The data URL representing the image. This property exists only if the request parameter
+ dataUrl String
imageType is 2 .

The pixel data in the pixels property represents a gray-scaled image captured by the camera of the HVC-P2. Each element in
the Array represents a pixel in the image. The value is an integer between 0 and 255 representing the gray-scale level. The first
element in the Array corresponds to the pixel of the top-left corner. The last element in the Array corresponds to the pixel of
the bottom-right corner. If the image size is 320x240, the number of the elements in the Array is 76800.

The value of the buffer property is a Buffer object representing the binary data of the image. The image format depends on
the parameter imageFormat .

The value of the dataUrl property is a data URL of the image like this:

data:image/png;base64,iVBORw0KGgoAAAANSU...AElFTkSuQmCC

Face detection

The code below sets all face detection options to '1' (enable) and creates an PNG file with markers in the current directory.

hvcp2.detect({
face: 1,
direction: 1,
age: 1,
 Page 9 of 21

gender: 1,
gaze: 1,
blink: 1,
expression: 1,
recognition: 1,
image: 1,
imageType: 3,
imageFormat: 'png',
imagePath: './capture.png',
imageMarker: true
});
}).then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The image below is the result of the detection. 3 faces are detected (2 human faces and 1 dog face). Note that the HVC-P2 is
arranged to detect human faces, not dogs.

(The original picture)

The result is as follows:

{
"face": [
{
"face": {
"x": 1062,
"y": 389,
"size": 252,
"confidence": 533
},
"direction": {
"yaw": 2,
"pitch": -15,
"roll": -2,
"confidence": 19
},
"age": {
"age": 67,
"confidence": 333
},
"gender": {
"gender": 1,
"confidence": 767
},
"gaze": {
"yaw": 3,
"pitch": -9
},
"blink": {
"left": 432,
"right": 407
},
"expression": {
"neutral": 1,
"happiness": 73,
"surprise": 0,
"anger": 8,
"sadness": 18,
"positive": 47
},
"recognition": {
 Page 10 of 21

"user": -127,
"score": -127
}
},
{
"face": {
"x": 728,
"y": 870,
"size": 240,
"confidence": 566
},
"direction": {
"yaw": 6,
"pitch": -13,
"roll": -12,
"confidence": 1
},
"age": {
"age": 45,
"confidence": 153
},
"gender": {
"gender": 1,
"confidence": 713
},
"gaze": {
"yaw": -4,
"pitch": 3
},
"blink": {
"left": 418,
"right": 465
},
"expression": {
"neutral": 44,
"happiness": 1,
"surprise": 15,
"anger": 39,
"sadness": 1,
"positive": -50
},
"recognition": {
"user": -127,
"score": -127
}
},
{
"face": {
"x": 618,
"y": 417,
"size": 216,
"confidence": 671
},
"direction": {
"yaw": 3,
"pitch": -20,
"roll": 1,
"confidence": 1
},
"age": {
"age": 68,
"confidence": 400
},
"gender": {
"gender": 0,
"confidence": 538
},
"gaze": {
"yaw": 4,
"pitch": -2
},
"blink": {
"left": 366,
"right": 455
},
"expression": {
"neutral": 0,
 Page 11 of 21

"happiness": 100,
"surprise": 0,
"anger": 0,
"sadness": 0,
"positive": 100
},
"recognition": {
"user": -127,
"score": -127
}
}
],
"image": {
"width": 320,
"height": 240
}
}

The meanings of the properties above is as follows:

Property Type Description

face Object The result of the face detection.

The x-coordinate of the center point of the detected face in the camera view port (1920 x
+x Number
1080)

The y-coordinate of the center point of the detected face in the camera view port (1920 x
+y Number
1080)

+ size Number The size of the square representing the detected face.

+ confidence Number The confidence in the face detection (0 - 1000).

direction Object The result of the face direction estimation.

The yaw angle of the detected face. The unit is degree. If the face looks to the left, this value
+ yaw Number
is positive. If the face looks to the right, this value is negative.

The pitch angle of the detected face. The unit is degree. If the face looks to the top, this value
+ pitch Number
is positive. If the face looks to the bottom, this value is negative.

The roll angle of the detected face. The unit is degree. If the face leans to the left, this value is
+ roll Number
positive. If the face leans to the right, the value is negative.

+ confidence Number The confidence in the face direction estimation (0 - 1000).

age Object The result of the age estimation.

+ age Number The age estimated from the detected face. The maximum value is 75.

+ confidence Number The confidence in the age estimation (0 - 1000).

gender Object The result of the gender estimation.

The gender estimated from the detected face. If the value is 0 , it means female. If the value
+ gender Number
is 1 , it means male.

+ confidence Number The confidence in the gender estimation (0 - 1000).

gaze Object The result of the gaze estimation.

The yaw angle of the gaze of the detected face. The unit is degree. If the eyes on the detected
+ yaw Number face look to the left, this value is positive. If the eyes on the detected face look to the right,
this value is negative.

The pitch angle of the gaze of the detected face. The unit is degree. If the eyes on the
+ pitch Number detected face look to the top, this value is positive. If the eyes on the detected face look to
the bottom, this value is negative.

blink Object The result of the blink estimation.

The blink degree of the left eye on the detected face. If the eye is fully opened, this value is 1.
+ left Number
If the eye is fully shut, this value is 1000.

+ right Number
 Page 12 of 21

Property Type Description

The blink degree of the right eye on the detected face. If the eye is fully opened, this value is
1. If the eye is fully shut, this value is 1000.

expression Object The result of the expression estimation.

The degree of the neutral estimated from the detected face. The range of value is from 0 to
+ neutral Number
100.

The degree of the happiness estimated from the detected face. The range of value is from 0
+ happiness Number
to 100.

The degree of the surprise estimated from the detected face. The range of value is from 0 to
+ surprise Number
100.

The degree of the anger estimated from the detected face. The range of value is from 0 to
+ anger Number
100.

The degree of the sadness estimated from the detected face. The range of value is from 0 to
+ sadness Number
100.

The degree of the positiveness estimated from the detected face. The range of value is from
+ positive Number -100 to 100. If the value is positive, the detected face looks like happy. If the value is negative,
the detected face looks like sad or angry.

recognition Object The result of the face recognition.

The user ID recognized from the detected face. If no user is recognized, this value will be
+ userId Number
-127.

The score of the recognition. The range of the value is from 0 to 1000. If no user is
+ score Number
recognized, this value will be -127.

Body detection

The code below sets the body detection option to '1' (enable) and creates an PNG file with markers in the current directory.

hvcp2.detect({
body: 1,
image: 1,
imageType: 3,
imageFormat: 'png',
imagePath: './capture.png',
imageMarker: true
});
}).then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The image below is the result of the detection. 3 bodies are detected.

(The original picture)

The result is as follows:

{
"body": [
{
"x": 1252,
 Page 13 of 21

"y": 674,
"size": 488,
"confidence": 889
},
{
"x": 780,
"y": 560,
"size": 480,
"confidence": 552
},
{
"x": 390,
"y": 656,
"size": 304,
"confidence": 642
}
],
"image": {
"width": 320,
"height": 240
}
}

The meanings of the properties above is as follows:

Property Type Description

The x-coordinate of the center point of the detected body in the camera view port (1920 x
x Number
1080)

The y-coordinate of the center point of the detected body in the camera view port (1920 x
y Number
1080)

size Number The size of the square representing the detected body.

confidence Number The confidence in the body detection (0 - 1000).

Hand detection

The code below sets the hand detection option to '1' (enable) and creates an PNG file with markers in the current directory.

hvcp2.detect({
hand: 1,
image: 1,
imageType: 3,
imageFormat: 'png',
imagePath: './capture.png',
imageMarker: true
});
}).then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The image below is the result of the detection. 2 hands are detected.

(The original picture)

The result is as follows:

 Page 14 of 21

{
"hand": [
{
"x": 1168,
"y": 699,
"size": 661,
"confidence": 1000
},
{
"x": 550,
"y": 675,
"size": 608,
"confidence": 972
}
],
"image": {
"width": 320,
"height": 240
}
}

The meanings of the properties above is as follows:

Property Type Description

The x-coordinate of the center point of the detected hand in the camera view port (1920 x
x Number
1080)

The y-coordinate of the center point of the detected hand in the camera view port (1920 x
y Number
1080)

size Number The size of the square representing the detected hand.

confidence Number The confidence in the hand detection (0 - 1000).

getConfigurations([params]) method
The getConfigurations() method reports the configurations of your HVC-P2. This method returns a Promise object. This
method takes a hash object as an argument containing properties as follows:

Property Type Required Description

true (default) or false . If true is specified or this property is not specified, this
cache Boolean Optional method reports the data cached when the connect() method was called. If false is
specified, this method requests to the HVC-P2.

The node-omron-hvc-p2 cashes the configurations when the connect() method is called. This method reports the cashed
configurations by default. If you want to make a query again, specify the cache property and set it to false .

The code blow calls this method and show the cached configurations.

hvcp2.getConfigurations().then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The code below calls this method and show the newly queried configurations.

hvcp2.getConfigurations({cache: false}).then((res) => {

console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The codes above will output the result as follows:

{
"cameraAngle": {
"angle": 0
 Page 15 of 21

},
"threshold": {
"body": 500,
"hand": 500,
"face": 500,
"recognition": 500
},
"detectionSize": {
"bodyMin": 30,
"bodyMax": 8192,
"handMin": 40,
"handMax": 8192,
"faceMin": 64,
"faceMax": 8192
},
"faceAngle": {
"yaw": 0,
"roll": 0
}
}

The meanings of the properties above are described in the next section " setConfigurations() method".

setConfigurations(configrations) method
The setConfigurations() method sets the configurations. This method returns a Promise object. This method takes a hash
object as an argument containing properties as follows:

Property Type Required Description

cameraAngle Object optional Camera Angle

+ angle Number optional 0: 0º, 1: 90º, 2: 180º, 3: 270º

threshold Object optional Threshold Values

+ body Number optional Human Body Detection threshold (1 - 1000)

+ hand Number optional Hand Detection threshold (1 - 1000)

+ face Number optional Face Detection threshold (1 - 1000)

+ recognition Number optional Face Recognition threshold (0 - 1000)

detectionSize Object optional Detection Size

+ bodyMin Number optional Human Body Detection minimum size (20 - 8192)

+ bodyMax Number optional Human Body Detection maximum size (20 - 8192)

+ handMin Number optional Hand Detection minimum size (20 - 8192)

+ handMax Number optional Hand Detection maximum size (20 - 8192)

+ faceMin Number optional Face Detection minimum size (20 - 8192)

+ faceMax Number optional Face Detection maximum size (20 - 8192)

faceAngle Object optional Face Angle Range

+ yaw Number optional Yaw angle range (0: ±30º, 1: ±60º, 2: ±90º)

+ roll Number optional Roll angle range (0: ±15º, 1: ±45º)

Though all properties are optional, at least one property must be specified.

See the "Human Vision Components (HVC-P2) B5T-007001 Command Manual (PDF)" for the details of the configurations.

The updated configurations will be pass to the resolve() function, which is a hash object containing the all properties as same
as the properties described above.

The code below set the camera angle to 270º.

hvcp2.setConfigurations({
cameraAngle: {
angle: 3 // Camera angle: 270º
 Page 16 of 21

}
}).then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The code above will output the result as follows:

{
"cameraAngle": {
"angle": 3
},
"threshold": {
"body": 500,
"hand": 500,
"face": 500,
"recognition": 500
},
"detectionSize": {
"bodyMin": 30,
"bodyMax": 8192,
"handMin": 40,
"handMax": 8192,
"faceMin": 64,
"faceMax": 8192
},
"faceAngle": {
"yaw": 0,
"roll": 0
}
}

resetConfigurations() method
The resetConfigurations() method restore the default settings. This method returns a Promise object.

The updated configurations will be pass to the resolve() function, which is a hash object containing the all properties as same
as the properties described above.

hvcp2.resetConfigurations().then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The code above will output the result as follows:

{
"cameraAngle": {
"angle": 0
},
"threshold": {
"body": 500,
"hand": 500,
"face": 500,
"recognition": 500
},
"detectionSize": {
"bodyMin": 30,
"bodyMax": 8192,
"handMin": 40,
"handMax": 8192,
"faceMin": 64,
"faceMax": 8192
},
"faceAngle": {
"yaw": 0,
"roll": 0
}
}
 Page 17 of 21

getFaceRecognitionUsers([params]) method
The getFaceRecognitionUsers() method reports a list of the user ID registered in the album. This method returns a Promise
object. This method takes a hash object as an argument containing properties as :

Property Type Required Description

true (default) or false . If true is specified or this property is not specified, this
cache Boolean Optional method reports the data cached when the connect() method was called. If false is
specified, this method requests to the HVC-P2.

The node-omron-hvc-p2 cashes the user ID registered in the album when the connect() method is called. This method
reports the cashed list. But if you want to make a query again, specify the cache property and set it to false .

The result will be pass to the resolve() function, which is a hash object containing the properties as follows:

Property Type Description

userIdList Array The list containing the user IDs registered in the album.

The code blow calls this method and show the cached information.

hvcp2.getFaceRecognitionUsers().then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The code below calls this method and show the newly queried information.

hvcp2.getFaceRecognitionUsers({cache: true}).then((res) => {

console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The codes above will output the result as follows:

{
"userIdList": [
1,
3
]
}

getFaceRecognitionData(params) method
The getFaceRecognitionData() method reports a list of the data ID associated with the specified user ID. This method returns a
Promise object. This method takes a hash object as an argument containing properties as follows:

Property Type Required Description

userId Number required The user ID. The value must be an integer between 0 and 99 .

If no data associated with the specified userId is registered in the HVC-P2, the reject() function will be called.

The code below get a list of data ID associated with the user ID 1 .

hvcp2.getFaceRecognitionData({
userId: 1
}).then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

The code above will output the result as follows:

 Page 18 of 21

{
"dataIdList": [
2
]
}

addFaceRecognitionData(params) method
The addFaceRecognitionData() method adds a face recognition data to the album in the HVC-P2. This method returns a
Promise object. This method takes a hash object as an argument containing properties as follows:

Property Type Required Description

userId Number required The user ID. The value must be an integer between 0 and 99 .

dataId Number required The data ID. The value must be an integer between 0 and 9 .

imageType Number optional 0 : Array (default), 1 : Buffer, 2 : Data URL, 3 : File

imageFormat String optional "gif" (default), "jpg" , or "png"

imagePath String conditional File path with file name (e.g., "/tmp/image.png" )

If a face is recognized, the result will be pass to the resolve() function, which is a hash object containing the properties as
follows:

Property Type Description

width Number The width of the registered face image. The value is always 64 .

height Number The height of the registered face image. The value is always 64 .

pixels Array The pixel data of the recognized face. This property exists only if the parameter imageType is 0 .

The Buffer object representing the recognized face image. This property exists only if the
buffer Buffer
parameter imageType is 1 .

The data URL representing the recognized face image. This property exists only if the parameter
dataUrl String
imageType is 2 .

Note that the target face has to be in the scope of the camera when this method is executed. If no face is recognized or 2 or
more faces are recognized, the reject() function will be called.

The code below adds a user with user ID 1 and data ID 2 and save the recognized face image as a file "face.png" .

hvcp2.addFaceRecognitionData({
userId: 1,
dataId: 2,
imageType: 3,
imageFormat: 'png',
imagePath: 'face.png'
}).then((res) => {
console.dir(res);
}).catch((error) => {
console.error(error);
});

The code above will output the result as follows:

{ width: 64, height: 64 }

The created PNG file is as follows:

(The original picture)

In order to accurize the face recognition, it is recommended to call this method some times for a face. The HVC-P2 can register
up to 9 data per user.
 Page 19 of 21

After adding face recognition data, try the detect() method with the parameter recognition as follows:

hvcp2.detect({
recognition: 1,
}).then((res) => {
console.log(JSON.stringify(res, null, ' '));
}).catch((error) => {
console.error(error);
});

If the registered user face is recognized, the result will be as follows:

{
"face": [
{
"recognition": {
"userId": 1,
"score": 988
}
}
]
}

deleteFaceRecognitionData(params) method
The deleteFaceRecognitionData() method deletes a face recognition data from the album in the HVC-P2. This method returns
a Promise object. This method takes a hash object as an argument containing properties as follows:

Property Type Required Description

userId Number required The user ID. The value must be an integer between 0 and 99 .

dataId Number required The data ID. The value must be an integer between 0 and 9 .

If no data corresponding to the specified userId and dataId is registered in the album, the reject() function will be called.

The code below deletes a face recognition data corresponding to the user ID 1 and the data ID 2 .

hvcp2.deleteFaceRecognitionData({
userId: 1,
dataId: 2
}).then(() => {
console.log('Deleted.');
}).catch((error) => {
console.error(error);
});

deleteFaceRecognitionUser(params) method
The deleteFaceRecognitionUser() method deletes all face recognition data associated with the specified user ID in the album.
This method returns a Promise object. This method takes a hash object as an argument containing properties as follows:

Property Type Required Description

userId Number required The user ID. The value must be an integer between 0 and 99 .

If no data associated with the specified userId is registered in the album, the reject() function will be called.

The code below deletes all face recognition data associated with the user ID 1 .

hvcp2.deleteFaceRecognitionUser({
userId: 1
}).then(() => {
console.log('Deleted.');
}).catch((error) => {
console.error(error);
});
 Page 20 of 21

clearFaceRecognitionData() method
The clearFaceRecognitionData() method deletes all face recognition data in the album. This method returns a Promise
object.

hvcp2.clearFaceRecognitionData().then(() => {
console.log('Deleted.');
}).catch((error) => {
console.error(error);
});

saveAlbum(params) method
The saveAlbum() method reports the album data set in the HVC-P2. It also can save the album data as a binary file. This
method returns a Promise object. This method takes a hash object as an argument containing properties as follow:

Property Type Required Description

path String optional File path with the file name.

If the path property is specified, this method will create a album file.

The result will be pass to the resolve() function, which is a hash object containing the properties as follows:

Property Type Description

album Buffer A Buffer object representing the album data.

The code below saves the album data as a file "HVCAlbum.hac" in the current directory.

hvcp2.saveAlbum({
path: 'HVCAlbum.hac'
}).then((res) => {
console.log(res);
}).catch((error) => {
console.error(error);
});

The code above will output the result as follows:

{ album: <Buffer 20 00 ...> }

The album property, whether or not the parameter path was passed to the saveAlbum() method, will be set.

loadAlbum(params) method
The loadAlbum() method load the album data to the HVC-P2. This method returns a Promise object. This method takes a
hash object as an argument containing properties as follows:

Property Type Required Description

buffer Buffer optional A Buffer object representing the album data.

path String optional A File path of the album data.

Though both the buffer and path are optional, one or the other must be specified.

The code below loads the album data saved in a album file "HVCAlbum.hac" which was saved by the saveAlbum() method.

hvcp2.loadAlbum({
path: 'HVCAlbum.hac'
}).then(() => {
console.log('Loaded.');
}).catch((error) => {
console.error(error);
});
 Page 21 of 21

saveAlbumOnFlashROM() method
The album data is saved in the volatile memory. Therefore, the album data will be gone if the HVC-P2 is turned off. If you want
to save the album data permanently (that is, save in the non-volatile memory), you can use the saveAlbumOnFlashROM()
method. This method returns a Promise object.

hvcp2.saveAlbumOnFlashROM();
}).then(() => {
console.log('Saved.');
}).catch((error) => {
console.error(error);
});

reformatFlashROM() method
The reformatFlashROM() method formats the non-volatile memory in the HVC-P2. This method returns a Promise object.

hvcp2.reformatFlashROM();
}).then(() => {
console.log('Saved.');
}).catch((error) => {
console.error(error);
});

Release Note
• v0.1.2 (2017-11-04)
◦ Improved the auto-detection of USB port on Linux.
• v0.1.1 (2017-09-23)
◦ Fixed a bug that an error was thrown when no device was found depending on a condition.
• v0.1.0 (2017-08-06)
◦ The auto detection of a USB serial port was Improved with more efficient way.
• v0.0.1 (2017-06-24)
◦ First public release

References
• OMRON B5T-007001 Human Vision Components (HVC-P2)
• Human Vision Components (HVC-P2) B5T-007001 Command Manual (PDF)
• OMRON's Image Sensing Technology - OKAO™ Vision
• OMRON SENSING EGG PROJECT

License
The MIT License (MIT)

Copyright (c) 2017 Futomi Hatano

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation
files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify,
merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT
LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.