Board surface defect detection plan

Project background

    The localization of continuous flat press machines has promoted the automation of artificial board production,but the surface defect detection at the end of the production line is stil manually identified with the naked eye,and this    detection method has many problems.On the one hand,prolonged exposure to fatigue with the naked eye can lead to    false positives and serious omissions,resuling in substandard quality of the factory's panels.Buyers may request returns    or compensation;On the other hand,at present,the continuous press production line has a speed of up to 1500mm/s.    When manually identifying with the naked eye,the speed of the board needs to be reduced to Omm/s,greatly reducing the    factory speed of particle board.In response to the drawbacks of low accuracy and low efficiency in manual inspection,due    to the increasing contradiction between the market demand for arificial board and the production quality and efficiency of    artificial board,in order to adapt to market changes,Artificial board production enterprises must comprehensively inspect    and control the surface qualty of artificial boards,in order to produce high-efficiency and high-precision artficial boards.  

    This scheme adopts automatic classification and detection equipment for wood panels,and automatically identifies surface defects of the panels through deep learning methods,overcoming the influence of fatigue and subjective factors in   manual identification.It achieves precise positioning of wood surface defects and improves the utilzation rate of the   panels.

Requirement Description

     In the  production  process  of  medium  and  high density  board(particleboard),there  are  defects  such  as large   shavings,glue   spots,oil   stains,debris,small    pits, and  sand   leakage  on  the   board   surface.Currently,it relies  on  visual  inspection  by  workers,which  leads  to problems such as high labor intensity,low efficiency,and low accuracy.

    Detection format:1.81X2.4M

    Detection defects: oil  stains,single-sided 

    Minimum detection accuracy:0.1mm

Solution

    This plan uses industrial vision devices combined with Al vision algorithms to detect surface defects on single  boards.The  equipment  is  installed  in  the  back  of  the  sanding  light,and  after  secondary  sanding,the positions are manually 

sorted.Using a single line scanning camera for single-sided detection.

Economic Benefit Analysis

■Increase  production  output  and  optimize  quality;

■Optimize  quality  to  eliminate  errors,and   reduce   customercomplaint  rates;

■ Can reduce the number of quality inspection personnel and flip board workers,etc.