Moving Target Recognition Based on HSV Color Space
Extracting of ROI (region of interest) has been widely used in the fields of detecting and tracking moving objects. HSV (hue saturation value) is a color space created according to the visual properties of colors, which can be used to describe colors quantitatively. Based on HSV, we calculate the co...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Office of Control and Information Technology
2020-01-01
|
| Series: | Kongzhi Yu Xinxi Jishu |
| Subjects: | |
| Online Access: | http://ctet.csrzic.com/thesisDetails#10.13889/j.issn.2096-5427.2020.01.600 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Extracting of ROI (region of interest) has been widely used in the fields of detecting and tracking moving objects. HSV (hue saturation value) is a color space created according to the visual properties of colors, which can be used to describe colors quantitatively. Based on HSV, we calculate the color information and recognize the color features in images. For the basic task of recognizing and capturing the armors mounted on the opponent robots in real-time in the RoboMaster National University Robot Competition, a detecting algorithm based on the color features was proposed. Firstly, we apply threshold segmentation to the source image, and calculate the shape parameters of the contours to find out all bright light bar in this image. Then HSV color space is used to judge the color of preliminary ROI to confirm whether it is our target or not. Before this procedure, we need analyze the pixels of the right colorful light bar to confirm a pixel range of HSV and RGB parameters. This range is also determined according to a sheet about corresponding pixel parameters of typical colors in HSV and RGB color space. The criterion to judge the right object is the percentage of the pixels whose HSV and RGB parameters are in the range we confirmed before. Experimental results show that the average processing speed can reach 102.33 frame per second and the recognition accuracy is about 97.4% by the proposed method. This algorithm has been successfully applied in the RoboMaster Robot Competition and the results show that robot can effectively track a moving robot in real-time. |
|---|---|
| ISSN: | 2096-5427 |