The time was 2007, when I purchased a desktop with 80GB HDD. Life was good. I had lots of space.
Google launched a new program named Google PhotoScan, specifically targeted towards ‘scanning’ photographs using a mobile camera.
I was wondering if the same technique can be used with a DSLR. Will it give better results? Recently I tried to ‘scan’ older photographs using my DSLR and I have decent results. I used the RAW mode as I found that tweaking the settings helped get a ‘scanned’ image that was matching the actual photograph in terms of colour.
Will taking multiple pictures of the same photograph and then stitching them back as a panorama give better results as compared to a single photograph?
For example, while taking a single picture, due to light reflection, I have encountered a glare in a region of the photo. By using multiple pictures, will I be able to eliminate / reduce this glare? I wonder.
I need to do a few experiments.
- Google PhotoScan, https://twentymegahertz.wordpress.com/2017/07/10/google-photoscan/
Google PhotoScan is an interesting Android application.
It claims to take pictures that are from the typical glare / reflection that we come across when trying to ‘scan’ a photograph (more so a glossy finish one) using a mobile.
To do this, it takes asks us to take multiple pictures of the photograph, which it then merges as a panorama. Very interesting technique.
I had tried it initially and got decent results. I need to try out the application now that it has undergone a few updates. I will post some samples in a few days.
In everyday life activities, it is challenging to create that one wow picture.
But that does not mean we have to stop trying.
The irony of sharing is that many people post pictures on social media for consumption by friends. They get many comments and likes.
Then you post a picture. And it gets a muted response.
Looking at the response, you wonder if the picture conveys the same message as the one you had in mind when you took the picture.
The best part of software development is that even individual programmers can develop world-class software using a single machine and their own experience and expertise, with little or no (real) outside help.
Recently, I read an article about the Google Pixel and its usage for night time photography. The author started with the statement that he taken a night time picture in USA using a DSLR, followed by shooting the same scene using a Google Pixel and the Google Nexus.
The author did multiple experiments and has detailed the steps taken.
While the images have come out very nice, the most important thing to note from the blog is that the author has taken multiple images using the phones, which were then combined to get a single image. If a similar technique had been used with a DSLR, the images would have been equally nice.
My point is simple. To make an comparison, ALL parameters in an experiment have to be same. Changing the parameters for one of the elements being compared is unfair.
- Experimental Nighttime Photography with Nexus and Pixel, https://research.googleblog.com/2017/04/experimental-nighttime-photography-with.html
Over a recent weekend, I took a bit too many photographs. I had around 560 (in a 16GB card), around 320 (in a 8GB card) and around 200 (in another 16GB card). So, that is nearly 1000 pictures.
The total number of files – RAW and JPG – combined was 2186 files. The size was 31.8 GB.
Now sorting the pictures is a headache.
Recently, I was watching a few videos on YouTube related to Apache NiFi. The videos were shot from a distance, due to which, a lot of the material being displayed / projected was not clearly visible. But that is not the point.
As the videos were shot from a distance, a few members of the audience were visible. Each time the presenter changed a slide, many of the participants took to their mobile phones and took a picture of the slide being displayed.