Page 408 - IJB-10-3

International Journal of Bioprinting In situ thermal monitoring in bioprinting

Figure 5. Steps of the procedure. Cropping (a) and rotation (b) and operations were applied to the VR images using MATLAB basic functions, to
standardize the application of segmentation algorithms. In the case of IR images, due to the off-axis positioning of the thermal camera, the roto translation
(c) operations were applied via MATLAB fitgeotform2d function only to the region of interest of the whole frame (d).

variations in object intensity. Adaptive thresholding, as built-in function. This method binarizes the image
opposed to global thresholding, offers a notable advantage using a locally adaptive threshold. The threshold
in scenarios where image illumination and contrast vary was automatically computed by adaptthresh Matlab
®
across different regions such as in our VR images, or built-in function based on the local mean intensity in
where there are variations in object intensity like in IR the neighborhood of each pixel at a given sensitivity
images, where these discrepancies are common. Adaptive s, which indicates sensitivity toward thresholding
thresholding addresses these challenges by determining a more pixels as foreground.
local threshold value for each pixel based on its immediate
neighborhood. This adaptive approach allows the algorithm (iii) The radius r and the sensitivity s within the algorithm
to account for local variations in intensity, resulting in were chosen following a visual inspection of the
improved segmentation accuracy and robustness. By quality of the segmentation process through an
tailoring the threshold to the specific characteristics of empirical approach. This led to r = 20 and s = 0.5.
our images, adaptive thresholding effectively mitigates the The proposed process monitoring approach relied on
shortcomings of global thresholding, making it a superior a comparison between both the binarized VR images and
choice for applications where precise object delineation IR images obtained with the respective nominal shape of
is paramount. each layer. The image registration with the nominal shape
The developed algorithm consisted briefly of two steps: was conducted with the utilization of the landmark points
previously mentioned. To accomplish this registration, we
(i) In the first step, the image underwent morphological employed the “fitgeotrans” function within the MATLAB
®
opening processing for removing small noises while software platform. This function was configured to utilize
preserving the shape and size of larger objects with the “nonreflectivesimilarity” property, a transformation
®
imopen and imsubtract Matlab built-in functions. type well-suited for preserving shape integrity while
With this operation, the image was eroded and enabling translation, rotation, and scaling adjustments.
then dilated using a disk structuring element of Notably, this approach ensured that the relative shapes
®
radius r with the strel Matlab built-in function for within the moving image remained unaltered, with the
both operations. primary variations being attributed to transformations
(ii) In the second step, a binary image was obtained from preserving parallelism and straightness, thereby upholding
the pre-processed thermal image by using Bradley’s the integrity of our comparative analysis. The comparison
“adaptive” method within the imbinarize Matlab evaluation was conducted by calculating the Dice
®
49
Volume 10 Issue 3 (2024) 400 doi: 10.36922/ijb.2021