



摘" 要:紅外熱成像測溫技術可以實現非接觸式測量,在帶式輸送機故障診斷中逐步應用,其優點是直觀、高效,且不受環境條件的影響,但紅外圖像在采集傳輸的過程中會受到多種噪聲的干擾,從而產生圖像模糊等問題,進而對后續故障診斷系統的圖像分割、特征提取等產生影響。為解決圖像模糊問題,研究幾種經典去噪的算法,提出一種改進閾值函數的小波閾值紅外圖像去噪算法,仿真實驗結果表明,相較于經典算法具有更好的去噪效果。
關鍵詞:紅外圖像;帶式輸送機;去噪算法;閾值函數;小波閾值
中圖分類號:TP391" " " 文獻標志碼:A" " " " " 文章編號:2095-2945(2024)26-0138-04
Abstract: Infrared thermal imaging temperature measurement technology can achieve non-contact measurement and is gradually used in belt conveyor fault diagnosis. Its advantages are intuitive, efficient, and not affected by environmental conditions. However, infrared images will be affected by the process of collection and transmission. The interference of various noises causes problems such as image blur, which in turn affects the image segmentation and feature extraction of subsequent fault diagnosis systems. In order to solve the image blur problem, several classic denoising algorithms were studied and an improved threshold was proposed. Functional wavelet threshold infrared image denoising algorithm, simulation experimental results show that it has better denoising effect than the classicalgorithm.
Keywords: infrared image; belt conveyor; denoising algorithm; threshold function; wavelet threshold
近年來,帶式輸送機在工業生產中扮演著越來越重要的角色,其應用領域涵蓋煤礦、礦山、港口等多個行業,如果發生故障,將會對整個線路造成影響,進而引發嚴重的事故,為減少帶式運輸機的故障率,紅外圖像診斷系統逐漸運用到帶式輸送機的故障診斷中[1],紅外圖像在采集、傳輸和存儲過程中,受外部環境以及電子元件等多種干擾因素的影響,會受到各種噪聲的干擾,其中高斯噪聲,椒鹽噪聲將嚴重影響紅外圖像的質量[2],進而影響到帶式輸送機紅外圖像的分割和識別,因此有必要將帶式輸送機紅外圖像進行去噪處理,有效的去噪算法是當前紅外圖像處理領域的熱點。
本文針對帶式運輸機紅外圖像的特點,對比分析幾種不同的去噪算法,其中就小波閾值去噪的閾值函數選取進行了詳細的分析,對閾值函數進行改進,提出了一種新的閾值函數,同時對閾值進行適當的調整,經過仿真驗證,去噪效果得到提升。……