The Inkjet Printing Image Processing Method Based on SRCNN Network
To address issues in inkjet printing images caused by the acquisition environment or system,such as image blurriness and insignificant parts of the image,we propose a method that utilizes grayscale image processing to set a three-dimensional topographic map and grayscale contour map based on the grayscale values of the image. This method aims to accurately determine the center position and edges of ink droplets in printed images,better distinguishing between overlapping and individual ink droplets. Additionally,we propose an inkjet printing image processing method based on the Super-Resolution Convolutional Neural Network (SRCNN) to enhance the clarity of low-resolution inkjet printing images. By adding ordinary convolution layers to the feature extraction layer to extract richer image features,this method has been validated on a standard dataset. The results in inkjet printing images show that the modified SRCNN model improves the peak signal-to-noise ratio by 0.014dB compared to the unmodified SRCNN model,demonstrating the effectiveness of the improved model and achieving better visual effects in reconstructed images.
万方数据
维普
