{"id":2765,"date":"2024-07-17T11:13:56","date_gmt":"2024-07-17T04:13:56","guid":{"rendered":"https:\/\/new-ocean.com.vn\/?post_type=article&p=2765"},"modified":"2024-08-14T16:18:39","modified_gmt":"2024-08-14T09:18:39","slug":"deep-learning-machine-vision","status":"publish","type":"article","link":"https:\/\/noas.new-ocean.com.vn\/en\/luu-tru\/article\/deep-learning-machine-vision","title":{"rendered":"Deep Learning and Machine Vision"},"content":{"rendered":"

\u201cN\u1ebfu b\u1ea1n \u0111ang m\u1eafc k\u1eb9t v\u1edbi c\u00e1ch l\u00e0m c\u0169 v\u00e0 kh\u00f4ng c\u00f3 kh\u1ea3 n\u0103ng s\u1ed1 h\u00f3a c\u00e1c quy tr\u00ecnh s\u1ea3n xu\u1ea5t, chi ph\u00ed c\u1ee7a b\u1ea1n c\u00f3 th\u1ec3 s\u1ebd t\u0103ng l\u00ean, s\u1ea3n ph\u1ea9m c\u1ee7a b\u1ea1n s\u1ebd \u0111\u01b0a ra th\u1ecb tr\u01b0\u1eddng mu\u1ed9n v\u00e0 kh\u1ea3 n\u0103ng cung c\u1ea5p gi\u00e1 tr\u1ecb gia t\u0103ng \u0111\u1eb7c bi\u1ec7t cho kh\u00e1ch h\u00e0ng s\u1ebd gi\u1ea3m,\u201d Stephen Ezell – chuy\u00ean gia v\u1ec1 ch\u00ednh s\u00e1ch \u0111\u1ed5i m\u1edbi to\u00e0n c\u1ea7u t\u1ea1i Qu\u1ef9 \u0110\u1ed5i m\u1edbi v\u00e0 C\u00f4ng ngh\u1ec7 Th\u00f4ng tin, cho bi\u1ebft trong m\u1ed9t b\u00e1o c\u00e1o c\u1ee7a Intel v\u1ec1 t\u01b0\u01a1ng lai c\u1ee7a AI trong s\u1ea3n xu\u1ea5t.<\/em><\/p>\n\n\n\n

In other words, companies that can quickly transform their factories into intelligent automation hubs will be the ones to gain long-term profits from those investments. These technologies have become essential to both manufacturing and business operations.<\/p>\n\n\n

\n
\"deep-learning-machine-vision\"<\/figure>\n<\/div>\n\n\n

According to a recent research report from Forbes Insights, 93% of survey respondents in the automotive and manufacturing sectors classified AI as \u201cextremely important\u201d or \u201cabsolutely essential for success.\u201d However, only 56% of these respondents plan to increase their AI spending\u2014and by just 10% or less of their current budgets. This hesitation to invest may stem from a lack of understanding about ROI and the practical applications of AI within their industries.<\/p>\n\n\n\n

So, how can AI be applied in manufacturing?<\/strong><\/mark><\/h2>\n\n\n\n

In visual inspection applications, AI-based image analysis specifically through Deep Learning<\/strong> and Machine Vision <\/strong>can significantly enhance factory efficiency. The combination of rule-based machine vision and deep learning\u2013based image analysis enables robotic assembly systems to precisely identify components, detect missing or misaligned parts, and quickly determine whether an issue exists.<\/p>\n\n\n\n

The synergy between machine vision and deep learning forms the foundation for manufacturers to implement smarter technologies, empowering them with greater scalability, accuracy, efficiency, and financial growth for the next generation. However, understanding the nuances between traditional machine vision and deep learning, and how they complement rather than replace each other, is essential to maximize the value of these investments.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

What is Deep Learning?<\/mark><\/h3>\n\n\n\n

Deep learning <\/strong>is a subset of artificial intelligence (AI) and a part of the broader field of Machine Learning.<\/strong> Instead of humans programming computer applications for specific tasks, deep learning uses data and trains it through neural networks to produce more accurate results based on that training data. Simply put, deep learning allows specific tasks to be solved without the need for explicit programming.<\/p>\n\n\n\n

Deep learning can consistently recognize anomalies and variations within a dataset on a large scale. This is something humans naturally do well \u2014 spotting differences \u2014 but something that traditional, rule-based computer systems have not been good at until now. However, unlike human inspectors, computers do not tire when making decisions on an assembly line.<\/p>\n\n\n\n

Applications of Deep Learning in Everyday Life:<\/strong> Facial recognition to unlock phones or identify friends in social media photos, recommendation engines on video and music streaming services or e-commerce websites, medical diagnosis, email spam filtering, and credit card fraud detection.<\/p>\n\n\n\n

\"nhan-dien-khuon-mat\"<\/figure>\n\n\n\n

How does a deep learning system work?<\/mark><\/h3>\n\n\n\n

According to O\u2019Reilly Media, there are five major categories of machine learning algorithms:<\/p>\n\n\n\n

Supervised learning<\/strong> involves mapping input data to known labels provided by humans. Recommendation tools used by online music and movie services apply supervised learning techniques.<\/p>\n\n\n\n

Unsupervised learning <\/strong>uses input data that isn\u2019t labeled, and the system tries to learn the structure of that data automatically without any human guidance. Anomaly detection \u2014 such as flagging unusual credit card transactions to prevent fraud \u2014 is an example of unsupervised learning.<\/p>\n\n\n\n

Semi-supervised learning<\/strong> is often a combination of the first two approaches. The system is trained on a partially labeled dataset \u2014 typically, a large amount of unlabeled data and a smaller portion of labeled data. Facial recognition in Facebook and Google\u2019s photo services are practical applications of this approach.<\/p>\n\n\n\n

Reinforcement learning <\/strong>is primarily a research field, but industrial use cases are beginning to emerge. It occurs when a computer system receives data within a specific environment and learns how to maximize outcomes based on that data. Google\u2019s DeepMind AlphaGo, which successfully learned how to play the board game Go, is a recent example of this technique.<\/p>\n\n\n\n

Transfer learning<\/strong> li\u00ean quan \u0111\u1ebfn vi\u1ec7c s\u1eed d\u1ee5ng l\u1ea1i m\u1ed9t m\u00f4 h\u00ecnh \u0111\u00e3 \u0111\u01b0\u1ee3c \u0111\u00e0o t\u1ea1o \u0111\u1ed3ng th\u1eddi gi\u1ea3i quy\u1ebft m\u1ed9t v\u1ea5n \u0111\u1ec1 v\u00e0 \u00e1p d\u1ee5ng v\u00e0o m\u1ed9t v\u1ea5n \u0111\u1ec1 kh\u00e1c nh\u01b0ng c\u00f3 li\u00ean quan. M\u1ed9t v\u00ed d\u1ee5 v\u1ec1 h\u1ecdc chuy\u1ec3n giao l\u00e0 m\u00f4 h\u00ecnh h\u1ecdc s\u00e2u \u0111\u01b0\u1ee3c \u0111\u00e0o t\u1ea1o tr\u00ean h\u00e0ng tri\u1ec7u h\u00ecnh \u1ea3nh c\u1ee7a m\u00e8o, sau \u0111\u00f3 \u0111\u01b0\u1ee3c “tinh ch\u1ec9nh” \u0111\u1ec3 ph\u00e1t hi\u1ec7n kh\u1ed1i u \u00e1c t\u00ednh trong ch\u1ee5p h\u00ecnh y t\u1ebf.<\/p>\n\n\n\n

\"deep-learning\"<\/figure>\n\n\n\n

What is Machine Vision?<\/mark><\/h3>\n\n\n\n

At its core, machine vision systems rely on digital sensors housed inside industrial cameras, equipped with specialized optics to capture images. These images are then sent to a computer (PC), where dedicated software processes, analyzes, and measures various characteristics to make decisions.<\/p>\n\n\n\n

Traditional machine vision systems perform reliably with consistent, well-manufactured parts. They operate through step-by-step filtering and rule-based algorithms, which are more cost-effective than large-scale human inspection. These systems can operate at extremely high speeds and with great precision \u2014 on production lines, rule-based machine vision systems can inspect hundreds or even thousands of parts per minute. The output of this visual data is based on a programmed, rule-based approach designed to solve inspection problems.<\/p>\n\n\n\n

In a factory setting, traditional rule-based machine vision is ideal for:<\/p>\n\n\n\n

Guidance: <\/strong>Locating, orienting, and identifying key features of a part to optimize other vision inspection tools in later stages.<\/p>\n\n\n\n

Identification: <\/strong>Reading barcodes, data matrix codes, direct part markings, and printed characters on parts, labels, and packaging.<\/p>\n\n\n\n

Measurement: <\/strong>Calculating distances between two or more points or geometric locations on an object and verifying whether these measurements meet specifications.<\/p>\n\n\n\n

Inspection: <\/strong>T\u00ecm c\u00e1c sai s\u00f3t ho\u1eb7c c\u00e1c b\u1ea5t th\u01b0\u1eddng kh\u00e1c nh\u01b0 kh\u00f4ng c\u00f3 ni\u00eam phong an to\u00e0n, c\u00e1c b\u1ed9 ph\u1eadn b\u1ecb h\u1ecfng, v.v…<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Comparison Between Deep Learning and Other Inspection Methods<\/mark><\/h3>\n\n\n\n

Deep Learning and Human Inspection<\/em><\/strong><\/p>\n\n\n\n

More Consistent:<\/strong> Reduces inconsistency between different human inspectors.<\/p>\n\n\n\n

More Reliable: <\/strong>Performs dependably even when scaled up or replicated across different production lines.<\/p>\n\n\n\n

Faster:<\/strong> Detects defects within milliseconds, enabling high-speed applications and improving throughput.<\/p>\n\n\n\n

Deep Learning and Traditional Machine Vision<\/em><\/strong><\/p>\n\n\n\n

Designed for Complex Applications: Solves challenging inspection and classification tasks that traditional rule-based algorithms cannot easily handle.<\/p>\n\n\n\n

Easier to Configure: <\/strong>Applications can be set up quickly, accelerating feasibility testing and project deployment.<\/p>\n\n\n\n

Accepts Variations: <\/strong>Handles defect variations for applications that require tolerance for deviations from strict control.<\/p>\n\n\n\n

High-Level Differences Between Traditional Machine Vision and Deep Learning<\/em><\/strong><\/p>\n\n\n\n

Development Process: Rule-based programming for each tool vs. example-based training.<\/p>\n\n\n\n

Hardware Investment: Deep learning requires greater processing power and storage capacity.<\/p>\n\n\n\n

Factory Automation Use Cases: Each approach is suited to different types of automation scenarios.<\/p>\n\n\n\n

\n
\n
\"\"<\/figure>\n<\/div>\n<\/div>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Conclusion<\/strong><\/mark><\/h2>\n\n\n\n

Rule-based machine vision and deep learning-based image analysis complement each other rather than serving as an either-or choice when implementing next-generation factory automation tools. In some applications, such as measurement, rule-based machine vision will continue to be the preferred and cost-effective option.<\/p>\n\n\n\n

For more complex inspections involving wide variations and unpredictable defects \u2014 too numerous and complicated to program and maintain in traditional machine vision systems \u2014 deep learning\u2013based tools provide an optimal and efficient alternative solution.<\/p>\n\n\n\n

Reference: Cognex<\/a><\/em><\/p>\n\n\n\n

>>> Read more:<\/strong> What Is a Vision System? <\/strong><\/a><\/p>\n\n\n\n

———————————————————- <\/p>\n\n\n\n

CONTACT INFORMATION<\/strong> <\/p>\n\n\n\n

New Ocean Automation System Company Limited <\/p>\n\n\n\n

Website:<\/strong> New Ocean Automation System<\/strong><\/a><\/p>\n\n\n\n

Hotline:<\/strong> 1900 0224 <\/p>\n\n\n\n

Email:<\/strong> customercare@new-ocean.com.vn<\/strong><\/a><\/p>","protected":false},"featured_media":2935,"template":"","article_cate":[14],"class_list":["post-2765","article","type-article","status-publish","has-post-thumbnail","hentry","article_cate-thong-tin-cong-nghe"],"acf":[],"_links":{"self":[{"href":"https:\/\/noas.new-ocean.com.vn\/en\/wp-json\/wp\/v2\/article\/2765","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/noas.new-ocean.com.vn\/en\/wp-json\/wp\/v2\/article"}],"about":[{"href":"https:\/\/noas.new-ocean.com.vn\/en\/wp-json\/wp\/v2\/types\/article"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/noas.new-ocean.com.vn\/en\/wp-json\/wp\/v2\/media\/2935"}],"wp:attachment":[{"href":"https:\/\/noas.new-ocean.com.vn\/en\/wp-json\/wp\/v2\/media?parent=2765"}],"wp:term":[{"taxonomy":"article_cate","embeddable":true,"href":"https:\/\/noas.new-ocean.com.vn\/en\/wp-json\/wp\/v2\/article_cate?post=2765"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}