The latest articles on DEV Community by Amaziah Yahalom (@amaziah_yahalom_de3c5fefc). https://dev.to/amaziah_yahalom_de3c5fefc https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3056638%2F21707fd6-74d5-4dd6-bbdd-985e923ec07d.png https://dev.to/amaziah_yahalom_de3c5fefc en Amaziah Yahalom Wed, 16 Apr 2025 19:35:08 +0000 https://dev.to/amaziah_yahalom_de3c5fefc/how-can-i-extract-and-interpret-charts-from-a-pdf-book-using-python-f1 https://dev.to/amaziah_yahalom_de3c5fefc/how-can-i-extract-and-interpret-charts-from-a-pdf-book-using-python-f1 <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fgt5qozdnmwjzxykuivez.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fgt5qozdnmwjzxykuivez.png" alt="Image description" width="800" height="641"></a></p> <p>I'm working on an AI trading assistant and have a specific challenge I'm hoping the dev and ML community can help with:</p> <p>I've loaded a full trading book into Python. The book contains numerous charts, figures, and graphs — like stock price plots labeled “FIGURE 104” with tickers like "U.S. STEEL". My goal is to extract these images, associate them with their captions (e.g., "FIGURE 104"), and generate meaningful descriptions or interpretations that I can feed into a reasoning AI model (I'm using something like DeepSeek locally).</p> <p>My question: 👉 What are the best Python tools or libraries for:</p> <p>Detecting and extracting images/figures from a PDF?</p> <p>Identifying chart features (e.g., axes, price levels, patterns)?</p> <p>Using OCR or other techniques to pull out relevant labels and text?</p> <p>Generating structured summaries that an AI model can reason over?</p> <p>Bonus: If you've done anything similar — like combining OpenCV, Tesseract, and a language model to describe visuals — I'd love to hear how you approached it.</p> programming python pdf