Closed amit2319 closed 5 years ago
4 is class 0.864626,0.946707,0.937679,0.855598,0.571571,0.515678,0.573691,0.629584, is the axis of four point of license plate after Normalizition
I want to map these 4 points on the input image of licence-plate-detection.py. Any help, I have tried x*width(img) and y*height(img) but points doesn't match the actual license plate.
@amit2319 these points are related to the cropped car image, not to the original image. You need to multiply by the width and height of the car bounding-box, and sum the top-left offset to retrieve the points in the original image space.
@sergiomsilva actually I was trying to retrive license plate from car_bounding_box only, but x.width(img) and y.height(img) where img refers to image of car_bounding-box doesn't reproduce the results.
In following data:
4,0.864626,0.946707,0.937679,0.855598,0.571571,0.515678,0.573691,0.629584,,
(x1,y1) = (0.864626,0.946707)
(x2,y2) = (0.937679,0.855598)
(x3,y3) = (0.571571,0.515678)
(x4,y4) = (0.573691,0.629584)
Am I interpreting it right?
Ah ok... now I understood. Its not the correct intepretation, considering your notation, the correct should be:
(x1,y1) = (0.864626, 0.571571) (x2,y2) = (0.946707, 0.515678) (x3,y3) = (0.937679, 0.573691) (x4,y4) = (0.855598, 0.629584)
In summary, the first 4 values are the x's, and the next 4, the y's.
@sergiomsilva Thanks! it worked.
licence-plate-detection.py returns *_lp.png* & _lp.txt**, can you describe data inside *_lp.txt.
sample data:
4,0.864626,0.946707,0.937679,0.855598,0.571571,0.515678,0.573691,0.629584,,