Combine state-of-the-art imaging and machine learning technology to accurately detect any kind of defect in wooden boards, ensuring a high-quality final product for your furniture.
Assure the inspection of every part with the most efficient and reliable defect detection in the production line. Take corrective measures upon any defect in real time.
Custom calibrations and out-of-the-box support of snapshot HSI cameras, allowing for a precise and reliable measurement of color in Lab coordinates. Unlike point colorimeters that measure the color on a spot, SpectralEdge allows the inspection of the whole board surface, detecting any local deviation from the specification.
Assure the absolute color of furniture parts, avoiding color drifts between batches that are hard to detect in production but become evident downstream, when parts from different batches are assembled.
Adapt quickly to production changes. Easily retrain inspection models to include references with new characteristics (e.g. new finishing or textures) and even new types of board defects.
SpectralEdge comes with a cloud-based API and easy to customize examples to build, store, and update datasets in the cloud and to train, validate, and deploy machine learning models back to the edge.
SpectralEdge enabled the implementation and deployment of a comprehensible and cost-effective solution to the monitoring needs of a leading manufacturer of wooden furniture for bathrooms. They have to inspect the whole production as furniture parts flow out of the production line. Defect detection at speed is a challenging and demanding task requiring several operators working in parallel. Given the production conditions, color measurement is performed offline on a reduced sample of parts.
Monitoring the production with a snapshot HSI camera combined with SpectralEdge has proven a solution to assure a complete inspection of every part in real time. Any defect or color deviation can now be detected and the corresponding part diverted for deciding about reworking or recycling the part.
A small dataset of images of wooden boards acquired with a selection of samples was enough to assess the feasibility of the inspection approach.
SpectralEdge accelerates and lowers the cost of development of solutions that assure quality inspection of furniture parts in real time in the production line.