Object Recognition and Moving Object Detection System Using Basic 2D Geometric Shape Approach and Accumulative Difference Image
As a part of computer vision, digital object recognition already ease many human activities which related with human vision. In this research, we used methods to recognize the car object and detect the moving object to help blind people. Image pre-processing, image segmentation, component (featur...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Subjects: | |
| Online Access: | https://digilib.itb.ac.id/gdl/view/33372 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | As a part of computer vision, digital object recognition already ease many human
activities which related with human vision. In this research, we used methods to
recognize the car object and detect the moving object to help blind people. Image
pre-processing, image segmentation, component (feature) extraction, confidence
rate measurement, balance point measurement, and pattern matching were the
stages to recognize the car object. The result by using this process gave 95% from
20 images successfully recognized under 500ms time elapsed. In this research
calibration stage also performed to measure the distance of moving object
detection using least square method. The Accumulative Difference Image (ADI)
method still gave good results although the image acquiring process have 150-
250ms latency (5 frames per second) for the object which move with maximum
speed 40km/h. The high latency or low frame per second acquiring rate caused
the object noise due movement sensitivity and also inaccurate area of moving
object. In order to detect the object with high speed (object with 80 km/h speed
rate) we use low latency acquiring rate 60-70ms (15 frames per second) to give
more accurate and safe results. |
|---|