查詢結果分析
相關文獻
- Directly Using RGB Color Space and Neural Network for Face Detection with Various Illumination
- 多階段人臉偵測與辨識
- Feature Matching Template for Detecting Eyes Open or Hidden from Web-Cam Images
- 模糊類神經網路於影像辨識之應用
- Neural Network Procedures for Taguchi's Dynamic Problems
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
- 類神經網路(Neural Networks)的種類及其在影像處理上的應用
- C++Fuzzy類神經網路物件導向發展系統之建立
頁籤選單縮合
題名 | Directly Using RGB Color Space and Neural Network for Face Detection with Various Illumination=直接使用RGB色彩空間與類神經網路做不同照度之臉部偵測 |
---|---|
作 者 | 林群雄; | 書刊名 | 統計與資訊評論 |
卷期 | 8 民94.12 |
頁次 | 頁43-65 |
分類號 | 312.1 |
關鍵詞 | 人臉偵測; RGB色彩空間; 色彩分割; 類神經網路; Face detection; RGB color space; Color segmentation; Neural networks; |
語文 | 英文(English) |
中文摘要 | 在本文中,我們建議的系統是由兩個主要部分組成。第一個部分是直接藉由皮膚顏色分割來二值化RGB 彩色影像。我們首先讀取一張RGB 彩色影像,藉由皮膚顏色分割,找出皮膚顏色區域。再將RGB 彩色影像直接轉變成二值化的影像(將皮膚顏色區域直接轉變成白色,其他直接轉變成黑色)。第二部分是先藉由等腰三角形為基礎來搜尋潛在臉的區域再以類神經網路完成臉部偵測的任務。我們首先藉由尋找等腰三角形的關係去得到潛在臉的區域,再將每一個潛在臉的區域先做尺寸標準化(60*60 畫素)的處理。然後,藉由類神經網路確認每一個人臉的正確位置。實驗結果顯示約百分之九十七的成功比率,並且相對錯誤比率很低。 |
英文摘要 | In this paper, we introduce a novel approach for automatic detection of human faces embedded in dissimilar lighting. The proposed system consists of two primary parts. The first part is to convert the input RGB color images to a binary image directly using RGB color segmentation. Since the absolute values of r, g, and b are totally different with the various skin colors and the altered lighting conditions, on the other hand, the relative value between r, g, and b are almost similar with the different skin colors and changed brightness circumstances. Therefore, we use the RGB color segmentation process to binarize the RGB color images directly instead of "color images to gray level images, then binary ones". For this reason, our system is very robust for different lighting conditions. We perform the human-skin color-segmentation task to remove the dissimilar lighting problem and got the binary images at the same time. The second part of the proposed system is to search the potential face regions and perform the task of face detection. In the second part, each face candidate is gotten from the isosceles-triangle criterion that is based on the rules of "the combination of two eyes and one mouth", and then to be normalized to a standard size (60*60 pixels). Next, each of these normalized potential face regions are fed to a neural network function to obtain the location of the face region. The proposed face detection system can detect color multiple faces embedded in dissimilar lighting conditions. Moreover, it can conquer different size, varying pose and expression. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。