Noctvision

Browse by topic

• OPTICS41 posts

  Optics

  This section covers the values that are settled by calculation first: the size of one pixel on the object, working distance (WD) and depth of field. How many pixels a defect occupies on screen is decided by these numbers before any algorithm runs.Start with theseHow to Calculate Machine-Vision Lens Focal Length — the basic formula from FOV, WD and sensor sizeMachine Vision Lens Calculator — where the shortcut formula breaks at higher magnification, plus depth of field and diffractionVision Types Split First by Wavelength and Pixel Size — several vision methods compared in one tableIn SWIR, One Stop Down Erases a 3.4 µm Pixel — how diffraction covers a pixel as wavelength growsUnfamiliar terms are explained in the Glossary, and a reading order across topics is in the Start Here guide.

  Start hereHow to Calculate Machine-Vision Lens Focal Length

• LIGHTING29 posts

  Lighting Design

  With the same camera, a defect can appear or vanish depending on the angle and wavelength of the light. This section collects articles on choosing among coaxial, dome, dark-field and backlight methods, and on handling reflection, wavelength and brightness drift.Start with theseSix Machine-Vision Lighting Methods and How to Choose — how each method works and the order for choosingFor Scratches on Highly Reflective Metal, Choose the Lighting First — a lighting map by surface typeWhy a Black Spot Remains in the Center of a Dome Light — the viewing-hole dark spot and fill lightWhen the Light Dims by 5 %, How Many Good Pixels Turn Into Defects? — how brightness change affects decisionsUnfamiliar terms are explained in the Glossary, and a reading order across topics is in the Start Here guide.

  Start hereSix Machine-Vision Lighting Methods and How to Choose

• ALGORITHM39 posts

  Vision Algorithm

  This section covers the rules and models that pick defects out of an image, and the cost when a decision is wrong: how rule-based and deep-learning methods share the work, anomaly detection, and filter order. It is the stage that turns the image made by lighting and optics into a pass/fail decision.Start with theseDeep Learning vs. Rule-Based Algorithms, When to Use Which — the conditions that favor eachOverdetection or Missed Detection, Which One Costs More? — setting the decision line by costBuilding a Vision Inspection App with PySide6 and OpenCV — full code for a minimal UI that tunes parameters on screenMust Every Normal Patch Be Remembered? — coreset design for memory-bank anomaly detectionUnfamiliar terms are explained in the Glossary, and a reading order across topics is in the Start Here guide.

  Start hereDeep Learning vs. Rule-Based Algorithms, When to Use Which

• EDGE AI22 posts

  Edge AI

  This section covers what goes wrong when an inspection model runs on a small board next to the machine instead of a server: accuracy lost to quantization, latency added by heat, and throughput set by the camera interface. These are items to confirm with numbers on site.Start with theseFour Snags in Initial Jetson Orin Nano Setup — a check order for first power-onIt Worked on the Server, So Why Does Accuracy Drop on the Edge? — the order for finding accuracy lossIs It Safe When the Average Latency Fits? — splitting the latency budget by p99Why Edge AI Vision Fails on the Factory Floor — a troubleshooting map from symptom to causeUnfamiliar terms are explained in the Glossary, and a reading order across topics is in the Start Here guide.

  Start hereIt Worked on the Server, So Why Does Accuracy Drop on the Edge?