Image detection project of the "bell-mark stamps", which is a traditional donation project in Japan. [2]

[The purpose] Detecting the characters on the image to divide each bell-mark stamp with the company's number, the number of stamps, and sum up the scores as the report. [Background] These stamps are applied to the product companies in Japan to support buying the school's equipment. Each school gathers these stamps with their effort and sums them up by sorting each company number and score. However, it takes too much time to conduct the work, and many volunteer parents try to do it manually. [Work package details] 1. Detect the numbers in each single image 2. Make the output list Each column includes the company's code, the number of stamps, and the sum of the score. We provide some of each single stamp for the detection The characters on each image: 1. Company's code e.x: ベルマーク番号 77 2. Score e.x: 1.5点 3. There are many sizes and colors, even in the same company, with the same score - The company code is always placed on top of the stamp. [Files] Please check the sample files on the project [W1] x 8 images And the others are here: cut the stamps only from the PDF. we only have an interest in the company code and the score. [login to view URL] [The environment] - Python - Works on Mac (Ventura), mac book air It is sufficient to work on the command line without the GUI in this version. - No need for additional graphic board [The deviverables] - Python code with library information - test output [Expected duration] 1week [Budget] - $30 (USD), however, we are preparing the additional payment in case the sophisticated worker Please do not hesitate to ask for the details. Best Regards,