Machine Vision System
A machine vision system (MVS) is a type of technology that enables a computing device to inspect, evaluate and identify still or moving images. These systems are used in manufacturing industries for inspection of random samples or in-line inspection of all samples.
A Machine Vision system typically consists of high-speed digital camera(s), Optics, Lighting system, Backend image processing and Decision-making system, PLC controls, Loading and Sorting mechanism.
We have rich experience in Vision systems and Automated inspection systems. We have a team of engineers with good experience in the field of Vision systems, Automation, Loading and Sorting mechanisms, Programming, Sensors/ InterfacesRobotics/Pick and Place systems, and Optics.
We use most advanced Imaging Software, which can do fast image processing, measurement/comparison, decision making and send triggers/result to other interfaces for appropriate actions. The software supports customization as per the application and also supports multiple cameras.
Some of our recent projects are
In-line inspection of Steel rods for Metal grading
This project was done for a Steel industry, manufacturing Steel pipes for Automotive industry. The goal was to do in-line inspection to check the Metal grade of each pipe coming out of the production line using a hand-held XRF analyser. Hand-held XRF analyser can do the analysis accurately with acceptable repeatability, if the sample is held close to the analyser head few millimetres. Hand held XRF analyser does the analysis of sample and gives and output on the result, which can be read into a PC. The analyser was positioned at the bottom side pointing upwards since there are samples with varying diameters.
The loading mechanism shall load the rod one at a time to a sliding mechanism which has an opening under which the XRF analyser head will be positioned. A stopper will be used to stop the rod at the end of the slider. Once the sample is stopped, a proximity sensor will sense the presence of the sample and will send a signal to trigger the XRF analyser. The analyser will analyse the sample and send the result to the PC. A software running on the PC will read the result and gives an OK or NOT-OK result. Based on the result, the sample shall be sorted into OK line or a bin to collect NOT-OK sample. The software also keeps a count of the samples – OK and NOT-OK count
Inspection system for checking Electrical panels
This system was developed for a customer manufacturing panels for electrical systems. This is an inspection system which checks for presence/absence of critical features like stamping, riveting and shape/bending. The customer has around 17 types of parts of different shape and size with different inspection requirement. The solution was a system with top camera and multiple lighting (top, and side). The software was taught about good samples and bad samples for all component types with multiple images of samples. The software will use this knowledge to give OK/NOT-OK result for the actual sample in production. The samples were loaded manually into the system and the inspection routine shall be triggered using sensor.
The camera will take multiple images of the sample using different lighting conditions. The images captured will be analysed and compared with the master data to decide whether the sample is OK or NOT-OK. If the sample is OK, the platform will be tilted to one side and the sample will slide gently into a conveyor from where it will be loaded for shipment. If the sample is NOT OK, then then platform will be tilted into another side and the sample will slide gently into a conveyor from where it will be collected into a bin. The system also keeps the statistics of OK/NOT-OK result. There is no measurement involved in this system, it only inspects the presence/absence of certain pre-defined features.