rollingthunder
Well-known member
Goldies.....
I'm not privvy to any sightings of the spp around here altho Craig regularly records individuals and small groups from his lofty perch @ Clent.
I have had to make do with counting a flock that was posted on LGRE's Tw@tter account taken at Slimbridge recently. It was part of a group of an estimated 5,000 birds, i would like to have seen the original with the estimate and work from there but i don't have the info/image to hand.
I initially did a visual estimage of maybe 750+ birds then i used the software. I did 4 different counts. First, open the image, then turn it into Grey-scale then use the Red slider to try and reduce each 'bird' to a Red Marker and finally make them into Black particles. At this last stage i then compare the original image with the particulated one and see whether it more or less matches the pattern of the flock and then at least you know you will be something near altho quite what the +/- factor is is anybodys guess?
At 130 it gave me 2317 dots/birds which i thought was way over and reset it at about 100 which gave 1671 which, again i thought an overestimate based on my visual first guess. A third setting of about 75 gave 1116 and finally a setting of about 60 gave 880 birds which is something more like it! The trouble with the last setting is that chunks of the flock dissappeared from the particulates so i knew that could'nt be right - frustating or what:C
Finally, i opened Photoshop CS6 Extended which has a 'count' tool. This is very useful as each click places a number on the bird. For this exercise i enlarged the image by 300%. There were still groups of upto a dozen to 15 that needed careful marking but i am confident there is very little missed.
To my surprise the individual click count showed a minimum of.....1696 birds:eek!: This is far more than i thought at the original image size so what do i know? It's an interesting excercise to carry out and i am sure that with practice on the software an optimum setting will be found in this case the slider set @ about 90/100 yielded the closest match.
Below are the original image size, the particulated image at same size and the enlarged one used on PS6 and click-counted (this took about 1/2 hour btw)
Laurie:t:
I'm not privvy to any sightings of the spp around here altho Craig regularly records individuals and small groups from his lofty perch @ Clent.
I have had to make do with counting a flock that was posted on LGRE's Tw@tter account taken at Slimbridge recently. It was part of a group of an estimated 5,000 birds, i would like to have seen the original with the estimate and work from there but i don't have the info/image to hand.
I initially did a visual estimage of maybe 750+ birds then i used the software. I did 4 different counts. First, open the image, then turn it into Grey-scale then use the Red slider to try and reduce each 'bird' to a Red Marker and finally make them into Black particles. At this last stage i then compare the original image with the particulated one and see whether it more or less matches the pattern of the flock and then at least you know you will be something near altho quite what the +/- factor is is anybodys guess?
At 130 it gave me 2317 dots/birds which i thought was way over and reset it at about 100 which gave 1671 which, again i thought an overestimate based on my visual first guess. A third setting of about 75 gave 1116 and finally a setting of about 60 gave 880 birds which is something more like it! The trouble with the last setting is that chunks of the flock dissappeared from the particulates so i knew that could'nt be right - frustating or what:C
Finally, i opened Photoshop CS6 Extended which has a 'count' tool. This is very useful as each click places a number on the bird. For this exercise i enlarged the image by 300%. There were still groups of upto a dozen to 15 that needed careful marking but i am confident there is very little missed.
To my surprise the individual click count showed a minimum of.....1696 birds:eek!: This is far more than i thought at the original image size so what do i know? It's an interesting excercise to carry out and i am sure that with practice on the software an optimum setting will be found in this case the slider set @ about 90/100 yielded the closest match.
Below are the original image size, the particulated image at same size and the enlarged one used on PS6 and click-counted (this took about 1/2 hour btw)
Laurie:t:
Attachments
Last edited: