No. SpotEgg was designed as a library of functions operated through a GUI which makes intensive use of Matlab Image Processing Toolbox, among others. If you have a valid matlab license, in addition to valid licenses for a number of toolboxes: Computer Vision System, Control System, Image Processing, Parallel Computing, Signal Processing, Statistics, then we can provide matlab .p files and run SpotEgg from Matlab prompt. However, if you don't have access to Matlab, we are also distributing a compiled version of SpotEgg wich should run flawlessly on Windows and MacOS. This compiled version runs a bit slower than the matlab desktop version but allows you to run SpotEgg even if you never used MATLAB before.
We now include a compiled version of dcraw together with SpotEgg. If you download it and compile it by yourself, please name your compiled file dcraw.exe under windows and dcraw under linux and move it to the same folder where you've uncompressed SpotEgg
Once you have downloaded and uncompressed the current realase of SpotEgg do as follows: 1) Start Matlab 2) In the navigation bar (on top of the command window) navigate to the folder where you uncompressed SpotEgg
3) In the command window just type SpotEgg and press enter. It should start.
First press CTRL-C to stop SpotEgg. Now open a file explorer window in your system and navigate to your RoIInfo folder under your RESULTS folder. There you'll find a file with the same name as the picture for which the incorrect RoI was accepted with the termination _VALID. Delete this file and run RoI definition again. When SpotEgg asks you whether to delete previous data click NO. Thus, you'll only run RoI definition again for the file you just deleted.
People use to believe it is possible while it is actually impossible. With two points we cannot find the photometric transfer curved implemented by your camera in the raw-to-jpeg transformation. SpotEgg requires a minimum of three points, and even in this case the accuracy of the fitting results is very poor. For jpeg files we recommend using at least six grey patches in the pictures.
Since release 1.4 SpotEgg offers the possibility of parallel processing for Spot Detection. Thus, different images are processed in parallel by different cores in your CPU. Modern CPUs offer at least four cores. For system reliability purposes we do not recommend to set the number of workers to the number of cores in your system as SpotEgg will take all the avaialble resources. Also take into account that running two workers requires twice as much memory as running a single one. Just for illustration purposes, we typically set this number to 2 in a 4 cores CPU.
When the user defines its input folder, SpotEgg runs a file name compatibility checking. If your file does not meet our requested name format, SpotEgg will automatically change file names accordingly to its needs. Please do not use your only available image folder for SpotEgg operation!!!, keep a backup of your files.
Currently SpotEgg accepts the following file formats for raw images: RW2, CR2, NEF, SRW, DNG. If this does not cover your needs, just contact us to find a solution for your file format. Sometimes adding a new format is extremely easy whereas other times it is a nightmare...I never know in advance. From time to time I try to add new formats support but I cannot match the pace of the industry.
Currently available version of SpotEgg accepts {.jpg .tiff .png .bmp} formats. But remember, your images need to include at least 3 different grey patches for algorithm convergence, and at least 6 different grey patches for good linearization accuracy.
SpotEgg main paper is :
Gómez, J. and Liñán-Cembrano, G. (2017), SpotEgg: an image-processing tool for automatised analysis of colouration and spottiness. J Avian Biol, 48: 502–512. doi:10.1111/jav.01117 Linearization uses D. Coffin DCRAW software during some parts of the process (to get raw data from your images) so we believe we also have to cite his work
Coffin, D. 2015. DCRAW application. – < www.cybercom.net/∼dcoffin/dcraw/ >
Yes, you have two options. 1) Start matlab and navigate to the folder where you saved the spot detection parameters file. In the matlab window, at the left, you'll see the Current Folder contents. Click on the name of the file. Just below this window, you'll see the content of the ParamSet vector. Copy this 4 values. Start SpotEgg and run spot detection. When it asks whether to load a file or train select train and do it manually. Now set the sliders to the values you just copied and modify them at your convenience. 2) Start matlab and navigate to the folder where you saved the spot detection parameters file. In the matlab window, at the left, you'll see the Current Folder contents. Drag and drop the file into the matlab command window. On top of the command window you'll see the WorkSpace. Doble click the ParamSet variable and modify the component you want to modify. Then, right click on the variable and select Save As. Overwrite the existing parameter file.
It is simple. First hit CTRL-C to stop SpotEgg. Now open an file explorer and navigate to the ChartInfo folder in your RESULTS folder. Delete the files XXXX_ORIGINAL.TIFF, XXXX_PATCHES.mat and XXXX_REFLECTANCE.TIFF, where XXXX is the name of the image file where the error was committed. Now start SpotEgg and rund Linearization again. When asked about deleting previous results click NO. Now SpotEgg should start at the first unprocessed image (the one you just deleted).
No. Flashes create, unless carefully employing diffusors, directional conditions for the illumination in your picture while it should be as diffuse as possible. You should always keep in mind that extracting reflectance values from a picture requires a number of assumptions.
(1) We need to assume that calibration target and subject receive the same amount of light power per area unit.
(2) We need to assume that they're both subtending the same angle from the camera optical axis
(3) We need to assume that both are lambertian surfaces (reflect equally in all directions) (4) We need to avoid shadows/reflections over subject and calibration target.
Yes, I see many images from researches using SpotEgg all around the world. Many times I see large images (>15Mpix) where the subjects cover less than 1% of the frame and spots are "pixelated" in a high resolution image. Always try to maximize the area your subject occupies in the image. Always try to optimize the framing by avoinding unnecessary items on the frame. Ideally, your subject and the calibration target (or chart) should occupy most of the scene. Additionally, if you're running a picture taking process over a large set of images, think before starting. Try to guarantee picture taking conditions that you can keep constant all accross the picture taking process. For instance, always try to have the target and the subject close to the center of the image. Try to maintain the distance between camera and objects approximately constant. Keep fixed ISO, focal length, illumination conditions, etc. for all the images. Picture taking is part of your science, be accurate there.
Don't panic. Most of times is a question of file/folder naming. What to check? 1) Verify that the whole path you introduced as input folder does not contain any space, or a character different from A-to-Z, a-to-z, numbers, - _ (If it does, rename these folders). 2) Verify that your image format is supported by SpotEgg. 3) Verify that your image files have the correct extension.
99% of the questions related to that are solved by following these 3 steps.