Classifynder Specification
(NB: Specifications and components subject to change depending on component availability).
Hardware
Power Supply
Input voltage (AC) range 100-240V, 50/60Hz range, 1.4A
Output voltage 24 volts, 3.75 Ampere, 90W
Stage performance-range, step resolution and scan areas
X – Y axis travel 136.5 x 36.5mm, minimum step 0.625µm (1600 steps per mm).
Z axis (focus controller) total travel 36.5mm, minimum step 0.312µm (3200steps per mm).
Slide area scanned: Rectangular or square scan area formats:
10 by 10mm, 22 by 22mm, 40 by 22mm, 48 by 20mm, 67 by 12mm
Manually selected region of interest scans may be made within a 67 by 22mm scan area window.
Microscopes and integral cameras
Low resolution:
Objective Magnification 4 x (~200 x effective at image formation due to digital magnification)
Monochrome CCD image sensor
Pixel size (on slide) 1.375 micrometer
Total number of pixels: 1280 (H) ◊ 1024(V) approx. 1.3M pixels
High resolution:
Objective Magnification 40x (~1000 x effective effective at image formation due to digital magnification)
(up to 2000 x possible with down-sampling turned off)
Monochrome CCD image sensor
Pixel size (on slide) 0.25 micrometer
Total number of pixels: 1280 (H) ◊ 1024 (V) approx. 1.3M pixels
Illumination Source
Custom array of high intensity Light Emitting Diodes (LEDs) at ~550nm (green)
Custom designed diffuser and stops for dark field illumination.
Communications
Serial interface USB 2
Cameras USB 2
Software
Autostage© release 3 or higher. Compiled L code written in National Instruments LabView.
Associated camera software drivers
User training video, guide and manual
Classification Accuracy:
The systems neural net classifier applied to images of undamaged acetolysed pollen has a classification accuracy between 97% and 80% depending on a range of factors, including sample preparation, diversity and the quality and representativeness of the image library used for training.
Particle size range:Between 5um and 100um
System Speed:
Low resolution scan speed is ~20 seconds per 10 square mm of sample area.
High resolution image capture and formation rate is 10 – 12 object images per minute.
Classification rate of finished high resolution images >100 images per second
Classifynder Shipping Components List
Classifynder generation 3B (Z axis dis-assembled for shipping)
Compact Disk (CD) or memory stick containing Classifynder software, drivers and web links to user training video, travel assembly/dis-assembly instructions and an operators reference guide (in English).
Modular 24 volt output mains power supply and lead (Country specific plug fittings available, US fitting supplied unless specified otherwise).
Two USB2 camera to PC connection leads.
Shipping and transport case
Shipping weight ~16.5Kgs
Items required but not supplied for the operation and use of the Classifynder
A Windows based laptop computer with Intel Core i7, 3GHz+ CPU, Windows 7 64bit OS Windows 7-64 bit, with at least 8GB RAM, three USB2 ports, a 16, 24, or 32-bit OpenGL capable graphics adapter.
A license for the Matlab© language version 7.5.0342 or above plus wavelet tool box and image processing toolbox.
A run time license for the LabView© NI Vision package.
Warranty
Comes with a 12 month return to base warranty.