Clojurians Log v2
introduce-yourself
Hi everyone, my name is Haider. I came across Clojure, when, two months ago, I decided to rewrite an open-source https://github.com/haiderriazkhan/CANDLE-J I had released a few years ago. To rewrite the program, I was looking for a JVM language that was functional and had robust built-in concurrency features. Clojure fit the bill, so I delved right in. It has quickly become my favourite programming language and I have found the community to be very kind and helpful. I’m looking forward to learning more about the Clojure ecosystem and meeting new folks.
Does the CANDLE algorithm require a 3D stack of images or does it also work with a stack of 2D images, e.g. from astronomical image aquisition?
Hi @lsolbach, CANDLE is specifically designed for denoising a 3D image stack—optimal for deep in vivo 3D multiphoton microscopy images where the signal to noise ratio is low. That said, the core component of CANDLE is the https://www.ipol.im/pub/art/2011/bcm_nlm/article.pdf, which is based on the premise that natural images have redundancies and that this self-similarity can be exploited to effectively denoise the image by averaging similar noisy patches present throughout the image. If that sounds applicable to your astronomical images, OpenCV has https://docs.opencv.org/3.4/d5/d69/tutorial_py_non_local_means.html of the non-local means filter for 2D images that you can use.