IMPLEMENTATION OF HAAR-FEATURE CASCADE CLASSIFIER ON PARKING LINE DETECTION AND TRACKING FOR VISION BASED PARKING ASSSIST SYSTEM (VPAS)
Parking line detection and movement tracking subsystem is one of the components that construct the Vision-Based Parking Assist System (VPAS). The subsystem used the HaarFeature Cascade Classifier as one of the methods to do the parking line detection. This model will identify the Haar-feature from...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/50570 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Parking line detection and movement tracking subsystem is one of the components that
construct the Vision-Based Parking Assist System (VPAS). The subsystem used the HaarFeature Cascade Classifier as one of the methods to do the parking line detection. This model
will identify the Haar-feature from an image, transform raw image to integral image,
implement the adaboost algorithm, and the cascade classifier. On the other hand, to tackle the
unstable classification process, tracking method will be implemented toward the system. The
parking line detection and movement tracking is implemented using the CPU with specification
of amd ryzen 3 2200g, motherboard msi a320 m pro vds, ram 4 gb ddr3, ssd 128gb, dan 4 cores
3.5 GHz processor. In result, the accuracy rate of right-side parking method is 0,87 while for
the left side is 0,877. Besides, the subsystem operates in 20-21 FPS.
|
|---|