Visão Computacional
O que é e qual sua importância?
Computer vision is a field of artificial intelligence that trains computers to interpret and understand the visual world. Using digital images from cameras and videos and deep learning models, machines can accurately identify and classify objects — and then react to what they “see.”
História da visão computacional
As primeiras experiências em visão computacional aconteceram nos anos 1950, com o uso de algumas das primeiras redes neurais para detectar os limites de um objeto e para classificar objetos simples em categorias como círculos e quadrados. Nos anos 1970, o primeiro uso comercial de visão computacional interpretou textos manuscritos e digitados usando reconhecimento ótico de caracteres. Esse avanço tinha como objetivo interpretar textos escritos para deficientes visuais.
Com o amadurecimento da internet nos anos 1990, grandes volumes de imagens foram disponibilizados online para análises, e o desenvolvimento de programas de reconhecimento facial explodiu. Esses crescentes conjuntos de dados ajudaram a possibilitar que máquinas identifiquem pessoas específicas em fotos e vídeos.
Hoje, inúmeros fatores convergiram para reanimar as pesquisas em visão computacional:
Tecnologias móveis com câmeras embutidas saturaram o mundo com fotos e vídeos.
O poder computacional tornou-se mais barato e compreensível.
Hardwares projetados para visão computacional e suas análises estão mais disponíveis.
Novos algoritmos como redes neurais convolucionais podem aproveitar melhor as capacidades dos hardwares e softwares.
Os efeitos desses avanços no campo da visão computacional têm sido surpreendentes. As taxas de precisão de identificação e classificação de objetos foram de 50 para 90% em menos de uma década — e os sistemas de hoje são ainda mais precisos que seres humanos nas rápidas detecção e reação a estímulos visuais.
Hear why Georgia-Pacific chose SAS
Computer vision resembles a jigsaw puzzle
Computers assemble visual images in the same way you might put together a jigsaw puzzle.
Think about how you approach a jigsaw puzzle. You have all these pieces, and you need to assemble them into an image. That’s how neural networks for computer vision work. They distinguish many different pieces of the image, they identify the edges and then model the subcomponents. Using filtering and a series of actions through deep network layers, they can piece all the parts of the image together, much like you would with a puzzle.
The computer isn’t given a final image on the top of a puzzle box — but is often fed hundreds or thousands of related images to train it to recognize specific objects.
Instead of training computers to look for whiskers, tails and pointy ears to recognize a cat, programmers upload millions of photos of cats, and then the model learns on its own the different features that make up a cat.
Visão computacional na atualidade
Do reconhecimento facial ao processamento das ações de um jogo de futebol ao vivo, a visão computacional concorre e supera as capacidades da visão humana em diversas áreas.
Deep learning e visão computacional
Como a tecnologia de deep learning ensina um computador a ver? Entenda como os diferentes tipos de redes neurais funcionam e como elas são usadas para a visão computacional.
Another set of eyes with computer vision
Georgia-Pacific embedded computer vision in day-to-day manufacturing operations to capture and analyze image data. By constantly monitoring for anomalies, the technology helps solve problems with quality and safety and boosts efficiency.
Análise de imagens e IA
Veja uma introdução à análise de imagens e aprenda técnicas analíticas que você pode aplicar aos dados de imagens.
Demonstração de reconhecimento facial
Aprenda as técnicas e etapas de processamento de dados subjacentes necessárias para o reconhecimento facial e a visão computacional. Essa demonstração expõe como o SAS® Viya® detecta, alinha, representa e classifica rostos.
Who's using computer vision?
Computer vision is used across industries to enhance the consumer experience, reduce costs and increase security.
Retail
Retailers can use computer vision to enhance the shopping experience, increase loss prevention and detect out-of-stock shelves. Computer vision is already helping customers checkout more quickly – aiding using self-checkout machines or combining with machine learning to alleviate the checkout process completely.
Manufacturing
In manufacturing, businesses use computer vision to identify product defects in real time. As the products are coming off the production line, a computer processes images or videos, and flags dozens of different types of defects — even on the smallest of products.
Government
Public Sector agencies use computer vision to better understand the physical condition of assets under their control, including equipment and infrastructure. Computer vision can help agencies perform predictive maintenance by analyzing equipment and infrastructure images to make better decisions on which of these require maintenance. In addition, Public Sector agencies use computer vision to help monitor compliance with policies and regulations. For example, computer vision can be used to detect contraband in cargo, flag potential safety violations in buildings, review labels for adherence to guidelines, and ensure compliance with conservation regulations. Finally, as drones become used more defense and homeland security needs, the use of analytics to identify and analyze critical elements from the visual feed will rise to the forefront of computer vision use cases in the public sector.
Health Care
In the medical field, computer vision systems thoroughly examine imagery from MRIs, CAT scans and X-rays to detect abnormalities as accurately as human doctors. Medical professionals also use neural networks on three-dimensional images like ultrasounds to detect visual differences in heartbeats and more.
Defense and Security
In high-security environments like banking and casinos, businesses use computer vision for more accurate identification of customers when large amounts of money are being exchanged. It’s impossible for security guards to analyze hundreds of video feeds at once, but a computer vision algorithm can.
Insurance
In the insurance industry, companies use computer vision to conduct more consistent and accurate vehicle damage assessments. The advancement is reducing fraud and streamlining the claims process.
Computer vision is one of the most remarkable things to come out of the deep learning and artificial intelligence world. The advancements that deep learning has contributed to the computer vision field have really set this field apart.
Learn about the multidisciplinary field of data science
Visão computacional é uma das coisas mais impressionantes gerada pelo mundo do deep learning e inteligência artificial. As contribuições que deep learning deu ao à visão computacional realmente destacou esse campo. Wayne Thompson Cientista de Dados do SAS
Visão computacional para conservação animal
Saiba como um modelo de visão computacional projetado para analisar rotas animais funciona. Um computador pode ser ensinado a enxergar uma pegada de modo similar a como um rastreador de animais veria? Veja como o computador processa as diferentes camadas de informação para determinar o animal e seu gênero. Nesse vídeo, Jared Peterson, Gerente Sênior de P&D de Inteligência Analítica Avançada do SAS, mostra como redes neurais são a grande ciência por trás da visão computacional.
Seeing results with computer vision
Computer vision users in many industries are seeing real results – and we’ve documented many of them in this infographic. For example, did you know:
- Computer vision can distinguish between staged and real auto damage?
- Computer vision enables facial recognition for security applications?
- Computer vision makes automatic checkout possible in modern retail stores.
From spotting defects in manufacturing to detecting early signs of plant disease in agriculture, computer vision is being used in more areas than you might expect.
Click on the infographic here to see results from retail, banking, health care and more.
How computer vision works
Computer vision works in three basic steps:
Acquiring an image
Images, even large sets, can be acquired in real-time through video, photos or 3D technology for analysis.
Processing the image
Deep learning models automate much of this process, but the models are often trained by first being fed thousands of labeled or pre-identified images.
Understanding the image
The final step is the interpretative step, where an object is identified or classified.
Today’s AI systems can go a step further and take actions based on an understanding of the image. There are many types of computer vision that are used in different ways:
- Image segmentation partitions an image into multiple regions or pieces to be examined separately.
- Object detection identifies a specific object in an image. Advanced object detection recognizes many objects in a single image: a football field, an offensive player, a defensive player, a ball and so on. These models use an X,Y coordinate to create a bounding box and identify everything inside the box.
- Facial recognition is an advanced type of object detection that not only recognizes a human face in an image, but identifies a specific individual.
- Edge detection is a technique used to identify the outside edge of an object or landscape to better identify what is in the image.
- Pattern detection is a process of recognizing repeated shapes, colors and other visual indicators in images.
- Image classification groups images into different categories.
- Feature matching is a type of pattern detection that matches similarities in images to help classify them.
Simple applications of computer vision may only use one of these techniques, but more advanced uses, like computer vision for self-driving cars, rely on multiple techniques to accomplish their goal.
Next Steps
Seeing is believing. Discover what SAS can help you achieve with computer vision.
Featured product for Computer Vision
SAS® Visual Data Mining and Machine Learning
Esta solução SAS suporta clustering, diferentes tipos de regressão, florestas aleatórias, modelos de aumento de gradiente, máquinas de vetores de suporte, análise de sentimentos e muito mais, além de deep learning. Um ambiente de pipeline interativo e visual apresenta cada projeto (ou meta) como uma série de etapas codificadas por cores que ocorrem em uma sequência lógica.
Leia também
- Artigo Analytics: A must-have tool for leading the fight on prescription and illicit drug addictionStates and MFCUs now have the analytics tools they need to change the trajectory of the opioid crisis by analyzing data and predicting trouble spots – whether in patients, prescribers, distributors or manufacturers. The OIG Toolkit with free SAS® programming code makes that possible.
- Artigo Inteligência artificial: mais perto de entender emoçõesHoje, existem dispositivos eletrônicos capazes de reconhecer vozes. Mas as pesquisas em IA apontam que o reconhecimento de emoções será possível em breve.
- Artigo What do drones, AI and proactive policing have in common?Law enforcement and public safety agencies must wrangle diverse data sets – such as data from drones – in their proactive policing operations. To be most effective, they need modern tools that support AI techniques like machine learning, computer vision and natural language processing.