jm

I want to create a program that assists with playing Gloomhaven (First Edition) in our board game club. The program should include the following features: 1. File Input (Rules & Data) Accept external files such as rulebooks, card data, or scenario descriptions. Use these files as reference material when analyzing the game state. Allow updates or new rules to be added later. 2. Card Recognition Use OCR (Optical Character Recognition) or image recognition to read cards from a camera or uploaded images. Identify abilities, numbers, and effects on the cards. Suggest possible strategies based on card combinations. 3. Board & Piece Recognition Use computer vision (e.g., OpenCV) to analyze live video or snapshots of the Gloomhaven board. Detect the hexagon grid, character minis, monster standees, and obstacles. Track their positions in real-time. 4. Movement & Attack Analysis Calculate optimal movement paths across hexes. Suggest possible attacks based on range, line of sight, area of effect, and monster positions. Follow the game rules from the rulebook when generating suggestions. 5. Interactive Assistant Behavior If the program is uncertain (e.g., about a rule interpretation or board state), it should ask the user to confirm or clarify. The assistant should not override the players but instead act as a helper. 6. Implementation Suggestions Computer Vision: - OpenCV (for hex detection, piece recognition). OCR: - Tesseract or similar (for card text). AI/Logic: - A rule-based engine combined with heuristics or AI for suggesting moves. User Interface: - A simple GUI or web-based dashboard where the video feed, recognized board state, and suggestions are displayed. The final goal is a digital game master assistant that: - Reads cards and rules. - Recognizes the board and pieces in real-time. - Calculates and suggests optimal moves and attacks. - Always confirms with the players before finalizing recommendations. The program should include these capabilities: 1. Input and Setup Accept external files (rulebook PDFs, scenario data, card lists). Let me draw or define the board boundaries manually if automatic hex detection is too complex. For example: Overlay a hex grid manually by drawing on top of the board image/video. Adjust hex size, rotation, and position to match the physical board. Allow me to label hexes (e.g., "obstacle", "trap", "starting position") by clicking or marking them. 2. Card Recognition & Suggestions Use OCR (like Tesseract) to read ability cards, monster cards, and other components from a camera or uploaded image. Store recognized cards in a structured format. Suggest strategies based on available player cards, discarded cards, and abilities. 3. Board and Piece Recognition Option A (automatic): Use computer vision (OpenCV) to detect minis/standees, hex grid, and obstacles from a live camera. Option B (manual assist): If the camera cannot detect pieces reliably, the program should allow me to click or drag pieces onto hexes in a digital overlay. This way, the program always knows the positions of characters and monsters, either automatically or with player help. 4. Movement & Attack Analysis Once the board and pieces are set up, the program should: Calculate all valid movement paths for a chosen character. Show which enemies are in range for melee or ranged attacks. Check line of sight and obstacles. Suggest optimal actions (movement + attack) according to the rules. The calculations should respect the official rulebook, which can be uploaded as reference. 5. Interactive Game Assistant If uncertain, the program asks the user for confirmation (e.g., "Is this hex an obstacle?" / "Is this ability card correct?"). Provide rule explanations on demand (like a teacher). Example: If a user tries an illegal move, the program explains why. If a user asks about “line of sight,” the assistant shows the rule and illustrates it. 6. Teaching Mode The assistant can explain step by step what is happening in the round. Example: “First, choose your two cards for the round.” “Now select the initiative order.” “Monsters draw their ability card next.” Highlight important reminders (e.g., “Don’t forget to reshuffle your modifier deck when the miss card is drawn”). 7. Implementation Details Core Technologies: - OpenCV for camera/board recognition. - Tesseract (or another OCR engine) for card text. - A rule engine (could be Python-based logic or AI heuristics). - GUI (Tkinter, PyQt, or a browser-based dashboard with Flask/React). Hybrid Mode: Support both automatic recognition (for speed) and manual override (for accuracy). Goal: - A digital Gloomhaven assistant that: - Reads rules & cards. - Lets me draw/setup the board manually if needed. - Recognizes the board and pieces (automatically or with player input). - Suggests moves & attacks. - Acts as a rule teacher and verifier. - Always asks me to confirm when it’s not sure