Quality Control in Industry with CV and TinyML - Revision history

Admino739mjm7: /* Acknowledgement */

2024-01-30T16:11:03Z

Acknowledgement

← Older revision		Revision as of 16:11, 30 January 2024
Line 77:		Line 77:
	==== Acknowledgement====		==== Acknowledgement====
	<p style="font-size:90%;line-height: 1.5em">		<p style="font-size:90%;line-height: 1.5em">
	''This work is an update of ExpertAI-Lux S.à r.l work in the context of the H2020 [https://www.airegio-project.eu/ AI REGIO project] (Grant Agreement No 952003). Further updates performed in the context of the AI REDGIO 5.0 project, concern; (1) leveraging tensorflow lite with the aim of being able to deploy on an edge device, and (2) post-training quantization with the aim of making the model more resource-efficient''		''This work is an update of ExpertAI-Lux S.à r.l work in the context of the H2020 [https://www.airegio-project.eu/ AI REGIO project] (Grant Agreement No 952003). Further updates performed in the context of the AI REDGIO 5.0 project, concern; (1) leveraging tensorflow lite with the aim of being able to deploy on an edge device, and (2) post-training quantization with the aim of making the model more resource-efficient.''
	</p>		</p>

	== Relevant Categories==		== Relevant Categories==
	[[Category:Jupyter Notebook]][[Category:Computer Vision]][[Category:Deep Learning]][[Category:Edge AI]][[Category:Quality Control]][[Category:Machine Vision]][[Category:Cloud-based]][[Category:Edge]]		[[Category:Jupyter Notebook]][[Category:Computer Vision]][[Category:Deep Learning]][[Category:Edge AI]][[Category:Quality Control]][[Category:Machine Vision]][[Category:Cloud-based]][[Category:Edge]]

Admino739mjm7: /* Relevant Categories */

2024-01-30T16:10:24Z

Relevant Categories

← Older revision		Revision as of 16:10, 30 January 2024
Line 81:		Line 81:

	== Relevant Categories==		== Relevant Categories==
	[[Category:~~Edge AI~~]][[Category:~~IoT~~]][[Category:~~Jupyter Notebook~~]][[Category:Edge]][[Category:~~Raspberry Pi~~]][[Category:~~Quality Control~~]][[Category:~~Computer Vision~~]]		[[Category:Jupyter Notebook]][[Category:Computer Vision]][[Category:Deep Learning]][[Category:Edge AI]][[Category:Quality Control]][[Category:Machine Vision]][[Category:Cloud-based]][[Category:Edge]]

Admino739mjm7 at 16:02, 30 January 2024

2024-01-30T16:02:31Z

← Older revision		Revision as of 16:02, 30 January 2024
Line 1:		Line 1:
	<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrolwithtinymlandcv.png\|thumb\|right\|<div style="font-size:88%;line-height: ~~2em~~">Image 1: The heatmap illustrates the contribution of each input image segment to each class. Segments with green colour contribute to the predicted class, while the red ones to the opposite class. <br>For example, if the model predicts that the given image is deficient: green parts were predicted as deficient, and the red ones as normal.<br> The third figure shows the area of the figure that mostly contributed to the predicted class</div>]]		<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrolwithtinymlandcv.png\|thumb\|right\|<div style="font-size:88%;line-height: 1.5em">Image 1: The heatmap illustrates the contribution of each input image segment to each class. Segments with green colour contribute to the predicted class, while the red ones to the opposite class. <br>For example, if the model predicts that the given image is deficient: green parts were predicted as deficient, and the red ones as normal.<br> The third figure shows the area of the figure that mostly contributed to the predicted class</div>]]

Admino739mjm7 at 16:02, 30 January 2024

2024-01-30T16:02:07Z

← Older revision		Revision as of 16:02, 30 January 2024
Line 1:		Line 1:
	<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrolwithtinymlandcv.png\|thumb\|right\|Image 1: The heatmap illustrates the contribution of each input image segment to each class. Segments with green colour contribute to the predicted class, while the red ones to the opposite class. <br>For example, if the model predicts that the given image is deficient: green parts were predicted as deficient, and the red ones as normal.<br> The third figure shows the area of the figure that mostly contributed to the predicted class]]		<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrolwithtinymlandcv.png\|thumb\|right\|<div style="font-size:88%;line-height: 2em">Image 1: The heatmap illustrates the contribution of each input image segment to each class. Segments with green colour contribute to the predicted class, while the red ones to the opposite class. <br>For example, if the model predicts that the given image is deficient: green parts were predicted as deficient, and the red ones as normal.<br> The third figure shows the area of the figure that mostly contributed to the predicted class</div>]]

Admino739mjm7 at 16:00, 30 January 2024

2024-01-30T16:00:46Z

← Older revision		Revision as of 16:00, 30 January 2024
Line 1:		Line 1:
	<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrolwithtinymlandcv.png\|thumb\|right\|Image 1: The heatmap illustrates the contribution of each input image segment to each class. ~~The segments~~ with green ~~color~~ contribute to the predicted class while the red ones to the opposite class. For example, if the model predicts that the given image is deficient~~. The~~ green parts ~~of the figure~~ were predicted as deficient, and the red ones as normal. The third figure shows the area of the figure that mostly contributed to the predicted class]]		<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrolwithtinymlandcv.png\|thumb\|right\|Image 1: The heatmap illustrates the contribution of each input image segment to each class. Segments with green colour contribute to the predicted class, while the red ones to the opposite class. <br>For example, if the model predicts that the given image is deficient: green parts were predicted as deficient, and the red ones as normal.<br> The third figure shows the area of the figure that mostly contributed to the predicted class]]

Admino739mjm7 at 15:59, 30 January 2024

2024-01-30T15:59:15Z

← Older revision		Revision as of 15:59, 30 January 2024
Line 1:		Line 1:
	<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:~~Qualitycontrol_snip~~.~~PNG~~\|thumb\|right\|Image 1: ~~Indicative output~~ of the ~~Quality Control notebook~~]]		<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrolwithtinymlandcv.png\|thumb\|right\|Image 1: The heatmap illustrates the contribution of each input image segment to each class. The segments with green color contribute to the predicted class while the red ones to the opposite class. For example, if the model predicts that the given image is deficient. The green parts of the figure were predicted as deficient, and the red ones as normal. The third figure shows the area of the figure that mostly contributed to the predicted class]]

Admino739mjm7 at 12:18, 30 January 2024

2024-01-30T12:18:47Z

← Older revision		Revision as of 12:18, 30 January 2024
Line 1:		Line 1:
	<strong>Quality control on production lines with computer vision and TinyML ~~for the automatic inspection~~ and ~~defect detection~~ in products as they move along the production line</strong> [[File:Qualitycontrol_snip.PNG\|thumb\|right\|Image 1: Indicative output of the Quality Control notebook]]		<strong>Quality control on production lines with computer vision and TinyML involves using computer vision algorithms and TinyML technology to automatically inspect and detect defects in products as they move along the production line</strong> [[File:Qualitycontrol_snip.PNG\|thumb\|right\|Image 1: Indicative output of the Quality Control notebook]]


Line 6:		Line 6:
	<p style="line-height: 1.5em">		<p style="line-height: 1.5em">

	'''Computer Vision (CV)''' ~~algorithms~~ can be used to ~~analyse~~ images of products and compare them to a "good" or "reference" image to identify any defects. These algorithms can be trained to detect a wide range of defects, including scratches, dents, misalignments, and missing components. Additionally, Computer Vision algorithms are able to work in real-time, allowing it to detect defects in products as they move along the production line and flag them for further inspection or rejection.		'''Computer Vision (CV)''' can be used to analyze images of products and compare them to a "good" or "reference" image to identify any defects. These algorithms can be trained to detect a wide range of defects, including scratches, dents, misalignments, and missing components. Additionally, Computer Vision algorithms are able to work in real-time, allowing it to detect defects in products as they move along the production line and flag them for further inspection or rejection.
	<br>		<br>
	</p>		</p>
	<p style="line-height: 1.5em">		<p style="line-height: 1.5em">
	'''TinyML''' is a field that involves developing machine learning models that can run on small, resource-constrained devices such as microcontrollers. In the context of quality control on production lines, TinyML can be used to enable the CV algorithms to run on embedded devices, such as cameras or sensors, that are integrated into the production line. This allows the system to process images and make decisions about defects without the need to send data to a separate computer for analysis.		'''TinyML''' is a field that involves developing machine learning models that can run on small, resource-constrained devices such as microcontrollers. In the context of quality control on production lines, TinyML can be used to enable the '''CV algorithms to run on embedded devices''', such as cameras or sensors, that are integrated into the production line. This allows the system to process images and make decisions about defects without the need to send data to a separate computer for analysis.
	<br>		<br>
	</p>		</p>
	<p style="line-height: 1.5em">		<p style="line-height: 1.5em">
	By combining computer vision algorithms and TinyML technology, it's possible to create a real-time, automated quality control system that		By combining computer vision algorithms and TinyML technology, it's possible to create a '''real-time, automated quality control system''' that can detect defects in products as they are produced, improving the overall quality and efficiency of the production process.
	can detect defects in products as they are produced, improving the overall quality and efficiency of the production process.
	</p>		</p>
	<br>		<br>
Line 25:		Line 24:
	! Feature !! Description		! Feature !! Description
	\|-		\|-
	\| Exploratory Data Analytics (EDA) on image data\|\| ~~''xxx''~~		\| Exploratory Data Analytics (EDA) on image data\|\| Pandas Dataframe, exporing and visualize the datasets and its metrics.
	\|-		\|-
	\| Data preparation and augmentation\|\| ''~~xxx~~''		\| Data preparation and augmentation\|\| Scale the values down to a smaller range, usually between 0 and 1. We use the '''tf.keras.layers.Rescaling''' layer in TensorFlow to accomplish this standardization and adjust the channel values of each image to be in the [0, 1] range. <br> Data augmentation increases the diversity of the training set by applying random (but realistic) transformations, such as image rotation.
	\|-		\|-
	\| Image classification with TensorFlow\|\| with Deep learning (CNN) models		\| Image classification with TensorFlow\|\| with Deep learning (CNN) models
	\|-		\|-
	\| Models evaluation\|\| ~~''xxx''~~		\| Models evaluation\|\| Confusion Matrix
	\|-		\|-
	\| Model interpretation predictions \|\| with LIME		\| Model interpretation predictions \|\| with LIME
Line 37:		Line 36:
	\| Transformation for use in embedded devices\|\| transformation to TFLite format		\| Transformation for use in embedded devices\|\| transformation to TFLite format
	\|-		\|-
	\| Post-training quantization \|\| ~~''xxx''~~		\| Post-training quantization \|\| Post-training quantization is a static quantization technique where the quantization is applied to the already trained model. This method can be applied to both weights and activations, or just the weights.
	\|-		\|-
	\| Quantization aware training\|\| '~~'xxx''~~		\| Quantization aware training\|\| Quantization aware training (QAT) involves training a model with knowledge of the quantization process that will be used to reduce the precision of the model's weights and activations during deployment.
	\|-		\|-
	\| Overall evaluation \|\| ~~''xxx''~~		\| Overall evaluation \|\| Evaluation based on metrics such as Accuracy, Recall, Precision, F1-Score, Inference Time and Size.
	\|-		\|-
	\| Real-time usage of the developed models \|\| with a python script ''opencv_object_tracking.py''		\| Real-time usage of the developed models \|\| with a python script ''opencv_object_tracking.py''
Line 54:		Line 53:
	* Model transformation to TinyML using TfLite		* Model transformation to TinyML using TfLite
	* Comparison between the original and TfLite mode.		* Comparison between the original and TfLite mode.
			</p>
			<br>

			<p style="line-height: 1.5em">
			The asset is Available ''(Note: For demonstration purposes, we use the casting product image data for quality inspection dataset available at [https://www.kaggle.com/datasets/ravirajsinh45/real-life-industrial-dataset-of-casting-product Kaggle]. However, a similar logic could be applied to other industries and product lines.)''
	</p>		</p>

Line 59:		Line 63:
	* Instructions available [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML?tab=readme-ov-file#instructions here]</li>		* Instructions available [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML?tab=readme-ov-file#instructions here]</li>
	* Detailed step-by-step demontration of the Quality Control asset can be found [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML/blob/main/Quality%20control%20on%20production%20lines%20with%20TinyML_v2.pdf here]</li>		* Detailed step-by-step demontration of the Quality Control asset can be found [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML/blob/main/Quality%20control%20on%20production%20lines%20with%20TinyML_v2.pdf here]</li>

			=== Licence===
			MIT License, licence link [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML/blob/main/LICENSE here]

	== Resources ==		== Resources ==
	<ul>		<ul>
	<li>Notebook available in [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML Github]</li>		<li>Notebook available in [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML Github]</li>
	<li>Created by [https://expertai-lux.com/ ExpertAI]</li>		<li>Created by [https://expertai-lux.com/ ExpertAI-Lux S.à r.l]</li>
	<li>Contact ~~''xxxx~~@~~xxx~~.~~xx''~~</li>		<li>Contact george.fatouros@expertai-lux.com / george.makridis@expertai-lux.com</li>
	</ul>		</ul>

Line 70:		Line 77:
	==== Acknowledgement====		==== Acknowledgement====
	<p style="font-size:90%;line-height: 1.5em">		<p style="font-size:90%;line-height: 1.5em">
	''This ~~repository was created~~ in the context of the H2020 [https://www.airegio-project.eu/ AI REGIO project] (~~AI REGIO is a project funded by the European Union Framework Programme for Research and Innovation Horizon 2020 under~~ Grant Agreement No 952003.)''		''This work is an update of ExpertAI-Lux S.à r.l work in the context of the H2020 [https://www.airegio-project.eu/ AI REGIO project] (Grant Agreement No 952003). Further updates performed in the context of the AI REDGIO 5.0 project, concern; (1) leveraging tensorflow lite with the aim of being able to deploy on an edge device, and (2) post-training quantization with the aim of making the model more resource-efficient''
	</p>		</p>

	== Relevant Categories==		== Relevant Categories==
	[[Category:Edge AI]][[Category:IoT]][[Category:Jupyter Notebook]][[Category:Edge]][[Category:Raspberry Pi]][[Category:Quality Control]][[Category:Computer Vision]]		[[Category:Edge AI]][[Category:IoT]][[Category:Jupyter Notebook]][[Category:Edge]][[Category:Raspberry Pi]][[Category:Quality Control]][[Category:Computer Vision]]

Admino739mjm7: /* User Journey */

2024-01-30T07:17:31Z

User Journey

← Older revision		Revision as of 07:17, 30 January 2024
Line 46:		Line 46:
	\|}		\|}

	== ~~User Journey~~ ==		== Usage ==
	<p style="line-height: 1.5em">		<p style="line-height: 1.5em">
	This Notebook assumes that there is available a dataset of labeled images of a product and demonstrates how the user can perform the following:		This Notebook assumes that there is available a dataset of labeled images of a product and demonstrates how the user can perform the following:
Line 56:		Line 56:
	</p>		</p>

	=== ~~Usage~~ Walkthroughs===		=== Walkthroughs===
	* Instructions available [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML?tab=readme-ov-file#instructions here]</li>		* Instructions available [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML?tab=readme-ov-file#instructions here]</li>
	* Detailed step-by-step demontration of the Quality Control asset can be found [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML/blob/main/Quality%20control%20on%20production%20lines%20with%20TinyML_v2.pdf here]</li>		* Detailed step-by-step demontration of the Quality Control asset can be found [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML/blob/main/Quality%20control%20on%20production%20lines%20with%20TinyML_v2.pdf here]</li>


	== Resources ==		== Resources ==

Admino739mjm7 at 11:53, 29 January 2024

2024-01-29T11:53:11Z

← Older revision		Revision as of 11:53, 29 January 2024
Line 1:		Line 1:
	<strong>Quality control on production lines with computer vision and TinyML for the automatic inspection and defect detection in products as they move along the production line</strong> [[File:~~Placehoder image~~.~~jpg~~\|thumb\|right\|Image 1: ~~Placeholder caption~~]]		<strong>Quality control on production lines with computer vision and TinyML for the automatic inspection and defect detection in products as they move along the production line</strong> [[File:Qualitycontrol_snip.PNG\|thumb\|right\|Image 1: Indicative output of the Quality Control notebook]]


Line 25:		Line 25:
	! Feature !! Description		! Feature !! Description
	\|-		\|-
	\| Exploratory Data Analytics (EDA) on image data\|\| xxx		\| Exploratory Data Analytics (EDA) on image data\|\| ''xxx''
	\|-		\|-
	\| Data preparation and augmentation\|\| xxx		\| Data preparation and augmentation\|\| ''xxx''
	\|-		\|-
	\| Image classification with TensorFlow\|\| with Deep learning (CNN) models		\| Image classification with TensorFlow\|\| with Deep learning (CNN) models
	\|-		\|-
	\| Models evaluation\|\| xxx		\| Models evaluation\|\| ''xxx''
	\|-		\|-
	\| Model interpretation predictions \|\| with LIME		\| Model interpretation predictions \|\| with LIME
Line 37:		Line 37:
	\| Transformation for use in embedded devices\|\| transformation to TFLite format		\| Transformation for use in embedded devices\|\| transformation to TFLite format
	\|-		\|-
	\| Post-training quantization \|\| xxx		\| Post-training quantization \|\| ''xxx''
	\|-		\|-
	\| Quantization aware training\|\| xxx		\| Quantization aware training\|\| ''xxx''
	\|-		\|-
	\| Overall evaluation \|\| xxx		\| Overall evaluation \|\| ''xxx''
	\|-		\|-
	\| Real-time usage of the developed models \|\| with a python script ''opencv_object_tracking.py''		\| Real-time usage of the developed models \|\| with a python script ''opencv_object_tracking.py''
Line 60:		Line 60:
	* Detailed step-by-step demontration of the Quality Control asset can be found [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML/blob/main/Quality%20control%20on%20production%20lines%20with%20TinyML_v2.pdf here]</li>		* Detailed step-by-step demontration of the Quality Control asset can be found [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML/blob/main/Quality%20control%20on%20production%20lines%20with%20TinyML_v2.pdf here]</li>

	~~==== Acknowledgement====~~
	~~<p style="font-size:90%;line-height: 1.5em">~~
	''This repository was created in the context of the H2020 [https://www.airegio-project.eu/ AI REGIO project] (AI REGIO is a project funded by the European Union Framework Programme for Research and Innovation Horizon 2020 under Grant Agreement No 952003.)''
	~~</p>~~

	== Resources ==		== Resources ==
Line 71:		Line 67:
	<li>Contact ''xxxx@xxx.xx''</li>		<li>Contact ''xxxx@xxx.xx''</li>
	</ul>		</ul>


			==== Acknowledgement====
			<p style="font-size:90%;line-height: 1.5em">
			''This repository was created in the context of the H2020 [https://www.airegio-project.eu/ AI REGIO project] (AI REGIO is a project funded by the European Union Framework Programme for Research and Innovation Horizon 2020 under Grant Agreement No 952003.)''
			</p>

	== Relevant Categories==		== Relevant Categories==
	[[Category:Edge AI]][[Category:IoT]][[Category:Jupyter Notebook]][[Category:Edge]][[Category:Raspberry Pi]][[Category:Quality Control]][[Category:Computer Vision]]		[[Category:Edge AI]][[Category:IoT]][[Category:Jupyter Notebook]][[Category:Edge]][[Category:Raspberry Pi]][[Category:Quality Control]][[Category:Computer Vision]]

Admino739mjm7: /* Resources */

2024-01-29T11:48:15Z

Resources

← Older revision		Revision as of 11:48, 29 January 2024
Line 68:		Line 68:
	<ul>		<ul>
	<li>Notebook available in [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML Github]</li>		<li>Notebook available in [https://github.com/giorgosfatouros/Quality-Control-in-Industry-with-CV-and-TinyML Github]</li>
	<li>Created by [https://expertai-lux.com/ ExpertAI ~~- SCCH~~]</li>		<li>Created by [https://expertai-lux.com/ ExpertAI]</li>
	<li>Contact ''xxxx@xxx.xx''</li>		<li>Contact ''xxxx@xxx.xx''</li>
	</ul>		</ul>