To address the issue of online monitoring of the effectiveness of laser delayering paint removal in localized or specific areas of aircraft skin, a machine vision-based method for online monitoring of the cleanliness of laser delayering paint removal is proposed. Using an established visual online monitoring platform, images of the laser paint removal process are acquired. First, algorithms such as Canny edge detection, dilation, and perspective transformation are applied to extract localized or specific areas of laser delayering paint removal. Next, the extracted images are converted from the RGB color space to the HSV color space. The OTSU algorithm is then used to segment the V channel image into different color paint layers. Subsequently, the ratio of the areas occupied by these layers is calculated using a binarization algorithm. Finally, the cleanliness of laser delayering paint removal is monitored, and the accuracy of the algorithm is verified. The results show that the machine vision-monitored cleanliness is 39.80%. Experiments demonstrate a 0.42% error between the theoretical and monitored area ratios, indicating high consistency. This method provides methodological guidance and theoretical support for online monitoring of laser paint removal on aircraft skin.