The paper introduces a novel technique called Joint Bilateral Upsampling (JBU) to enhance image analysis and processing tasks such as tone mapping, colorization, stereo depth, and photomontage. These tasks often require computing solutions at a lower resolution due to computational and memory constraints, but traditional upsampling methods can introduce blurring and artifacts. JBU leverages the high-resolution input image as a prior to improve the upsampling process, resulting in better high-resolution solutions. The technique applies a spatial filter to the low-resolution solution and a range filter to the high-resolution image, effectively preserving edges and discontinuities. The paper demonstrates the effectiveness of JBU through various applications, showing that it outperforms traditional upsampling methods in terms of visual and numerical accuracy. The complexity of JBU is \(O(Nr^2)\), where \(N\) is the output image size and \(r\) is the domain filter radius, making it efficient for large images. The authors conclude that JBU is a promising approach for improving the quality of upsampling in image processing tasks.The paper introduces a novel technique called Joint Bilateral Upsampling (JBU) to enhance image analysis and processing tasks such as tone mapping, colorization, stereo depth, and photomontage. These tasks often require computing solutions at a lower resolution due to computational and memory constraints, but traditional upsampling methods can introduce blurring and artifacts. JBU leverages the high-resolution input image as a prior to improve the upsampling process, resulting in better high-resolution solutions. The technique applies a spatial filter to the low-resolution solution and a range filter to the high-resolution image, effectively preserving edges and discontinuities. The paper demonstrates the effectiveness of JBU through various applications, showing that it outperforms traditional upsampling methods in terms of visual and numerical accuracy. The complexity of JBU is \(O(Nr^2)\), where \(N\) is the output image size and \(r\) is the domain filter radius, making it efficient for large images. The authors conclude that JBU is a promising approach for improving the quality of upsampling in image processing tasks.