Detection of humans in input videos and drawing of rectangular frames from detected points on humans.

Board: MacnicaHelioSoCEvaluationKit
Tools Version: 13.1
State: running
Members: dynasystem

Video

Abstract

In this demonstration the target for detection is human upper body in input videos and detected humans are framed with red rectangles.

Software embedded in ARM core detects upper body of humans with maximum detection points of 10 from which the logic in FPGA draws rectangles that frame detected humans.

Video input/output is via DVI with resolution of 720p (60fps).

Detection accuracy will be improved for product use in accordance with each application.

Development Environment
Category Environment
OS Windows7 Professional SP1 64bit
Synthesis Altera Quartus2 v13.1.4(SP4)
Main Board Macnica HelioBoard rev1.3
ARM ARM DS-5 v5.15.0
Daughter Card Bitec HSMC-DVI
Video Input/Output 1280x720 Progressive 60fps
Hardware Composition

Following is the hardware composition used for this demonstration.

env.JPG

env2.JPG


Operational procedure
  1. Write sdcardimg.img in SD card in advance.
  2. Write demo.jic file in FPGA.
  3. Turn on main board. DVI input signals will be output to the display as they are.
  4. When "login" is displayed on PC console, input "root".
  5. Input "./main" to start Human Recognition application.
  6. Human upper body is detected and a red rectangular frame will be displayed.
  7. Push any key to quit the application.
*Maximum duration of this demonstration is up to 5 minutes.

© 1999-2024 RocketBoards.org by the contributing authors. All material on this collaboration platform is the property of the contributing authors.

Privacy Policy - Terms Of Use

This website is using cookies. More info. That's Fine