Larry Roberts was an American engineer who received the Draper Prize in 2001 “for developing the Internet“.
Computer Vision is the computer science field of detecting, categorizing, and quantifying information from visual data. If AI is the broad field of replicating the human mind using computers, Computer Vision is the replication of the visual cortex using layers of algorithms to replicate the parts of our brains that we take for granted. Fields like Artificial Neural Networks, Machine Learning, and deep learning are used today to analyze complex data patterns, make predictions, and derive business insights.
Visual data is far more chaotic than you would think. The human brain’s visual cortex does a lot of work under the surface to add layers of description on top of the shades and colors our eyes sense.
Before you can recognize your cat, your brain has to convert the orange blob surrounded by dark green and blue into lines that form edges, gradients that imply volume, and comparisons against your field of view that imply physical size.
Computer Vision: A Timeline
In the mid-20th century, computer vision emerged from the curiosity of scientists like David Marr and Larry Roberts, who laid the foundational principles for machines to process visual data. Over time, advancements in algorithms, hardware capabilities, and data availability set the evolution of computer vision in motion.
In 1963, Larry Roberts detailed the process of extracting 3-D information from 2-D photographs. In 1966, Marvin Minksy directed a graduate student to connect a camera to a computer and tasked it with describing its visual observations. Edge detection, feature extraction, and shape analysis also gained popularity, allowing computers to perform easy tasks like object recognition and scene understanding. Developing algorithms and robust dedicated hardware helped bring the dream of artificial sight closer to reality.
Researchers discovered that computers could learn to recognize complex patterns and objects accurately. This paved the way for new applications, from facial recognition systems to self-driving vehicles.
The past decade has witnessed a sea change in computer vision with deep learning – a subfield of machine learning that uses neural networks with multiple layers. Layers refer to the different levels of computation within the network. Each layer learns from the previous layer, performs some processing, and passes the result to the next layer. These layers are arranged hierarchically, with each subsequent layer learning more complex features or patterns from the input data. Think of it like peeling an onion: each layer uncovers more information and builds upon the previous layer’s understanding, ultimately leading to a deeper and more nuanced data comprehension.
Today, computer vision infiltrates virtually every aspect of our lives, from smartphone cameras automatically tag photos to surveillance systems monitoring public spaces. In healthcare, computer vision aids in medical imaging and diagnostics, while in manufacturing, it streamlines quality control processes.
AI and Computer Vision
AI emerged simultaneously as computer vision with early efforts to give machines human-like cognitive abilities. The term “artificial intelligence” was coined in the 1950s, sparking a wave of research aimed at creating machines capable of reasoning, learning, and problem-solving. While computer vision focused on replicating human vision, AI sought to replicate a broader range of human functions like language understanding, decision-making, and creativity. Over time, advances in AI have intersected with the development of computer vision, leading to breakthroughs and pushing both fields forward in tandem.
Marvin Lee Minski was an American cognitive and computer scientist mainly concerned with research of artificial intelligence (AI)
Applications for Computer Vision
Computer vision has many use cases in today’s businesses and industries.
Manufacturing:
Quality Control: Computer vision identifies flaws in products on assembly lines.
Robotic Automation: Robots are guided in various tasks including picking and placing items, assembly, and quality inspection.
Predictive Maintenance: Using computer vision technology to monitor machinery enables predictive maintenance, forecasting when maintenance is needed to reduce downtime and maintenance costs.
Agriculture:
Precision Agriculture: Combining computer vision with drones allows for the analysis of crop health, detection of pests and diseases, and optimization of irrigation and fertilizer usage, resulting in increased crop yields and reduced environmental impact.
Livestock Monitoring: Computer vision facilitates tracking livestock health and behavior, enabling early disease detection and efficient management.
Computer Vision and Plainsight
At Plainsight, we integrate computer vision and artificial intelligence to Filter data for your business. Plainsight Vision Intelligence Filters automate data gathering, inform decision-making, and allow your business to extract actionable insights from images and videos. By using AI-driven Filters, companies can access computer vision easily, reshaping how they use visual data and changing how they do business.
Conclusion
With Plainsight Filters, computer vision becomes a passive observer and an active participant in understanding and interpreting the visual world. Whether identifying objects, recognizing patterns, or extracting meaningful information from images and videos, Plainsight Filters empower computer vision systems to see beyond the surface and uncover more profound insights. Insights, of course, look very different for every business. Plainsight Filters use the combined power of computer vision and artificial intelligence to reveal hidden objects and allow for decision-making and success.
Learn more about Plainsight Filters and request a demo.
