Machine vision on mobile devices

Mobile devices (smartphones, tablets,…), especially last generation are replacing computers, because they are technologically in constant growth, have more processing capacity and lets execute more complex applications.

In addition the huge advantages introduced with touch screens, different sensors and embedded functionality (accelerometer, GPS, phone,…) expands the possibilities of mobile applications.

From INVAR aware of this reality during last months we are working in development of new computer vision applications in this type of devices.

Therefore, in order to bring these kind of applications to any person or company interested in the subject, We send it to you for free. Contact us and request it filling this form.


App setup

This application uses OpenCV libraries so you have to install previously in your device the OpenCV Manager that you can download directly from the oficial page of Google Play Store.


App description

When application starts, a real time image of the camera device is displayed. Clicking on the screen a menu bar appears and allows access to different options.

In next picture there is an example where you can see that the menu bar appears at the bottom of the screen.


Option. Face detection

¿What is it?  In real-time the application looks for facial patterns indicating his location inside the image by a green box

Try  Focus with your camera to someone that you have around and you will see how detects the faces. In the screen will appear the face with the green box. In the next image, you can see an example about a image record directly in the screen of the computer.


Uses   Video surveillance applications, people counting and recognition, …

Option. Edge detection

¿What is it?  Detect areas of the image with different contrast, allowing the identification of object’s shape.

Try   Focus your camera around and see the effect of this filter. What do you think?

In the following pictures you can see some examples of this filter on the dossier and the logo images of our company.

deteccion_bordes1 detección_bordes2 detección_bordes3 deteccion_bordes4

Uses   Searching and recognition of objects or certain features in images.

Option. RGB – HSV histogram

¿What is it?  Images are composed of pixels with information of the intensity of color contained in each of the three channels: Red (R), green (G) and blue (B).

RGB color space is the most widely used on image sensors (cameras), but it is also possible to represent the color in other spaces such as HSV corresponding to H (Hue), S (Saturation) and V (Value -> Brightness or intensity).

RGBRGB color space

HSV HSV color space

Histograms represents color intensity on a scale of 0 to 255 (256 values ​​- 8 bits).

This application displays in real time and from left to right the representation of this six (6) histograms (R, G, B, H, S, V).


Try  To see how this feature works you can focus with the camera an object for example of red color  and you will see how the first histogram corresponding to the channel ‘R’ ​​shows more values concentrated on the right side of the graph. That means the sensor is detecting intensities close to pure red (R = 255 value). In the histogram of the hue ‘H’  you will see as more values ​​are displayed in red color. The same happens to the other channels, green (G) and blue (B) where these colors dominate.


Uses   The information registered by the histograms is useful in inspection processes and quality control.

Option. Points of interest

¿What is it?   A point of interest in an image is a pixel that is in an area where there is a high contrast to the surrounding or adjacent pixels.

Try  Look around with the camera and you will see how some points of interest are marked with a small blue circle.

In the example image you can see that the highlighted points of interest are identified in blue and it’s possible to identify some features of the object.


Uses   3D vision, image restitution, patterns matching, shapes, …

Option. Zoom

¿What is it?   Display at the top left part of the screen what you can see in the central red box.

Try  Focus anywhere and you will see how it expands the middle of the image.

The following image shows an example of an electronic device where you can see with more detail some components.



Comments are closed.