I need to extract text segments from a grid image for OCR. I have tried multiple approaches such as HoughLines, Connected Components, Morpological operations etc. but I haven't got satisfactory results. Could anyone suggest a better approach? I have attached a few sample images