ObjectFinder

Version 9.1 (10 Mo) par Luca Della Santina
Recognize 3D structures in volumetric images
5.0
(2)
761 téléchargements
Mise à jour 5 fév. 2024
Afficher la licence sur GitHub

ObjectFinder is a MATLAB app that allows you to recognize a large number of small structures within a three-dimensional image volume.

This app is developed for neuroscience research, with the purpose of detecting fluorescently-labeled synapses in neuronal image stacks acquired using confocal or super-resolution microscopes.

Key features:
- Multi-threaded 3D object connectivity search within large image volumes
- Trainable deep learning classifier for automatic validation of objects
- Visual interaction with objects using the builtin volume inspector
- 3D inspection and interaction of detected objects using Bitplane Imaris
- Automated colocalization analysis
- Automated nearest neighbor analysis
- Integrated plots of detected object's statistics
- Export analysis results to Microsoft Excel®
- Batch processing of multiple images with custom start time

For more information and to download latest ObjectFinder version visit: https://lucadellasantina.github.io/ObjectFinder/

Citation pour cette source

Luca Della Santina (2024). ObjectFinder (https://github.com/lucadellasantina/ObjectFinder), GitHub. Récupéré le .

Compatibilité avec les versions de MATLAB
Créé avec R2021a
Compatible avec les versions R2021a et ultérieures
Plateformes compatibles
Windows macOS Linux
Catégories
En savoir plus sur Microscopy dans Help Center et MATLAB Answers

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

ImarisXTensions

Les versions qui utilisent la branche GitHub par défaut ne peuvent pas être téléchargées

Version Publié le Notes de version
9.1

New volume inspector. New skeleton management system. Bugfixes
8.8

+ Compiled binaries for Windows, macOS, Linux
+ Redesigned website
+ Calculation of object density by depth along Z-axis (whole-mount) and Y-axis (sections)

8.7

Batch reporting of object properties
8.1

Improved user interface, bugfixes
8.0

Simplified GUI, Multiple skeleton support, Visual object refinement
7.8

Faster loading times for large objects sets
Improved batch processing
New project button
7.5.2

Speed improvements and bugfixes
7.4

Faster loading speed
Improved reporting
Improved colocalization analysis
7.2

New Monte Carlo simulations
Improved deep-learning training
Support for image files larger than 4Gb
Bugfixes
7.0

Import and use custom Neural Network models from ONNX, Keras-Tensorflow or Caffe formats
New Colocalization mode: engulfed objects
New faster data storage for objects storage

6.5

Batch colocalization results can be saved into a table
Fixed error when closing manual colocalization analysis window
If no Z-resolution is present in TIF file, use ImageJ default of '1'
6.4

Automation improvements: Batch processing of all folders contained within a root path & batch colocalization analysis
6.2

+ Overwrite objects if already present
+ Allow filtering of objects based on their Z position
6.1

User can choose whether to use local or global noise detection
New purge invalid objects button allows to save on disk only validated objects
6.0

Improved detection algorithm
Simplifies user interface
Bug fixes
5.18

Improved Colocalization reports
Improved image selection dialog
Simplified user interface to detect objects
5.17

Integrated object inspector: improved accuracy of object selection and handling of images with non-square image ratio
5.16

User can now visually inspect non-colocalized objects
5.14

Fixed missing update of colocalization lists when a new experiment folder is loaded
5.13

Linear density along skeleton is also reported by depth
Volume occupancy of objects is also reported within mask
During image selection mask=0 always means no mask
Fixed bug in automatic colocalization analysis with binary mask
5.11

Support for 2D images
Fixed error when image files are missing voxel resolution
5.10

Support for skeletons created with ImageJ's Simple Neurite Tracer plugin
5.9

+ Improved speed when found objects are > 1 million
+ Added compatilibity with MATLAB R2018b
+ Added tooltips for all major UI components
5.8

Improved search accuracy by lowering stepping of intensity values within the volume to the finest value
Added convenience buttons in "About" tab with links to Home page, report a bug and user manual
5.7

Improved speed of blocks conflict resolution by ~200X
No more need to delete empty objects after resolution
5.6

Improved speed of objects accumulation by 10X
Improved speed of blocks conflict resolution by ~40X
Reduced block overlap needed for computation
5.5

Improved 10 times the speed of search algorithms by code optimization
Fixed bug when opening image files of different sizes
5.4

User can select among different search methods
New default search method "Local thresholding" up to 40X faster than previous default "Iterative thresholding"
Simplified definition of minimum and maximum object size range
5.3

Simplified resolution of duplicated objects across overlapping regions between blocks
Fixed insufficient block buffer and size calculation
Remember previous search settings when creating a new set of objects
5.2.1

Save colocalized objects as a new set of objects
Bugfixes
5.2

8 new deep learning models available for automatic objects validation (vgg-16/19, SqueezeNet, GoogleNet, Inception-v3, Resnet-50 / 101, Inception-ResNet-v2)

You can filter objects based on their shape properties (roundness and major axis length)
5.1

Improved 5X the speed of objects' validation when using neural network
5.0.1

Minimum requirements update
5.0

- Machine learning classifier using MATLAB's Neural Networks toolbox
- Plotting statistics of objects such as size / brightness / shape distribution
- Multiple sets of objects now allowed within the same experiment
- Nearest neighbor analysis
4.10.0.0

Colocalization analysis
Volume inspector
4.7.0.0

Linked to GitHub repository
4.6.0.0

Minor bugfix
- Slice inspector for online inspection and filtering of found objects
- Automate tab allows batch processing of multiple images
- Reduced memory usage per multi-threading worker
- Improved processing speed by ~30%
New icon
4.4.0.0

Project description update for v4.4
Added direct link to Matlab App installer
Description update

Pour consulter ou signaler des problèmes liés à ce module complémentaire GitHub, accédez au dépôt GitHub.
Pour consulter ou signaler des problèmes liés à ce module complémentaire GitHub, accédez au dépôt GitHub.