|
1 | | -from . import load_data |
| 1 | +import time |
2 | 2 |
|
3 | | -VALUES = load_data.load_values() |
4 | | -TEMPLATES = load_data.load_templates() |
| 3 | +import cv2 |
| 4 | +import mouse |
| 5 | +import numpy as np |
| 6 | +from PIL import ImageGrab |
| 7 | + |
| 8 | +from .load_data import load_values, load_templates |
| 9 | + |
| 10 | + |
| 11 | +class Vision: |
| 12 | + def __init__(self, screen_res: tuple[int, int]): |
| 13 | + self.screen_res = screen_res |
| 14 | + self.values = load_values() |
| 15 | + self.cv_templates = load_templates(self.screen_res) |
| 16 | + |
| 17 | + @staticmethod |
| 18 | + def color_distance(color1: tuple[int, int, int], color2: tuple[int, int, int]) -> float: |
| 19 | + """ Returns the distance between two colors in the RGB color space.""" |
| 20 | + return np.linalg.norm(np.array(color1) - np.array(color2)) |
| 21 | + |
| 22 | + def line_length(self) -> int: |
| 23 | + """ Returns the length of the line in the game or -1 if the line length is not detected.""" |
| 24 | + |
| 25 | + # grab the part of the screen where the line length is displayed |
| 26 | + top_left = self.values["line_length_pos"][f"{self.screen_res[0]}x{self.screen_res[1]}"]["top_left"] |
| 27 | + bottom_right = self.values["line_length_pos"][f"{self.screen_res[0]}x{self.screen_res[1]}"]["bottom_right"] |
| 28 | + bbox = (top_left[0], top_left[1], bottom_right[0], bottom_right[1]) |
| 29 | + cv_img = cv2.cvtColor(np.array(ImageGrab.grab(bbox=bbox)), cv2.COLOR_RGB2GRAY) |
| 30 | + |
| 31 | + # dictionary of x positions and the detected digits and their match values at that position |
| 32 | + # x_pos_digit_map = {x_position: [(digit, match_value), ...], ...} |
| 33 | + x_pos_digit_map = dict() |
| 34 | + digits_variations = len(self.cv_templates["digits"]) // 10 |
| 35 | + for i in range(len(self.cv_templates["digits"])): |
| 36 | + result = cv2.matchTemplate(cv_img, self.cv_templates["digits"][i], cv2.TM_SQDIFF) |
| 37 | + positions = np.where(result <= 11_500_000) |
| 38 | + positions_x = positions[1] |
| 39 | + positions_y = positions[0] |
| 40 | + |
| 41 | + # dictionary of x positions and the minimum match value at that position for current digit |
| 42 | + x_pos_matches = dict() |
| 43 | + for x, y in zip(positions_x, positions_y): |
| 44 | + if x in x_pos_matches.keys(): |
| 45 | + x_pos_matches[x] = min(x_pos_matches[x], result[y, x]) |
| 46 | + else: |
| 47 | + x_pos_matches[x] = result[y, x] |
| 48 | + |
| 49 | + # add the minimum value for each x position to the x_pos_digit_map |
| 50 | + for x in x_pos_matches.keys(): |
| 51 | + if x in x_pos_digit_map.keys(): |
| 52 | + x_pos_digit_map[x].append((i // digits_variations, x_pos_matches[x])) |
| 53 | + else: |
| 54 | + x_pos_digit_map[x] = [(i // digits_variations, x_pos_matches[x])] |
| 55 | + |
| 56 | + # normalize the x positions (merge the x positions that are close to each other) |
| 57 | + coordinates = list(x_pos_digit_map.keys()) |
| 58 | + norm_x_pos_digit_map = dict() |
| 59 | + while len(coordinates) != 0: |
| 60 | + coord = coordinates.pop() |
| 61 | + coord_values = x_pos_digit_map[coord] |
| 62 | + for i in range(-10, 11): |
| 63 | + if coord + i in coordinates: |
| 64 | + coord_values.extend(x_pos_digit_map[coord + i]) |
| 65 | + coordinates.remove(coord + i) |
| 66 | + norm_x_pos_digit_map[coord] = coord_values |
| 67 | + |
| 68 | + x_pos_ordered = list(norm_x_pos_digit_map.keys()) |
| 69 | + x_pos_ordered.sort() |
| 70 | + line_length = "" |
| 71 | + for i in x_pos_ordered: |
| 72 | + digit = min(norm_x_pos_digit_map[i], key=lambda x: x[1])[0] |
| 73 | + line_length += str(digit) |
| 74 | + |
| 75 | + # return the line length as an integer or -1 if the line length is not detected |
| 76 | + if line_length == "": |
| 77 | + return -1 |
| 78 | + else: |
| 79 | + return int(line_length) |
| 80 | + |
| 81 | + def fish_caught(self) -> bool: |
| 82 | + """ Check if the fish is caught (and process the pop-ups), return True if the fish was caught, False otherwise.""" |
| 83 | + img = cv2.cvtColor(np.array(ImageGrab.grab()), cv2.COLOR_RGB2GRAY) |
| 84 | + |
| 85 | + disc = cv2.minMaxLoc(cv2.matchTemplate(img, self.cv_templates["fish"]["discard"], cv2.TM_SQDIFF)) |
| 86 | + if disc[0] <= 1000000: |
| 87 | + mouse.move(disc[2][0], disc[2][1], absolute=True, duration=0) |
| 88 | + mouse.click(button="left") |
| 89 | + time.sleep(2) |
| 90 | + return True |
| 91 | + |
| 92 | + keep = cv2.minMaxLoc(cv2.matchTemplate(img, self.cv_templates["fish"]["keep"], cv2.TM_SQDIFF)) |
| 93 | + if keep[0] <= 1000000: |
| 94 | + mouse.move(keep[2][0], keep[2][1], absolute=True, duration=0) |
| 95 | + mouse.click(button="left") |
| 96 | + time.sleep(2) |
| 97 | + return True |
| 98 | + |
| 99 | + rel = cv2.minMaxLoc(cv2.matchTemplate(img, self.cv_templates["fish"]["release"], cv2.TM_SQDIFF)) |
| 100 | + if rel[0] <= 1000000: |
| 101 | + mouse.move(rel[2][0], rel[2][1], absolute=True, duration=0) |
| 102 | + mouse.click(button="left") |
| 103 | + time.sleep(2) |
| 104 | + return True |
| 105 | + |
| 106 | + return False |
| 107 | + |
| 108 | + def fish_hooked(self) -> bool: |
| 109 | + # grab the part of the screen where the fish hooked bar is displayed |
| 110 | + top_left = self.values["fish_hooked_pos"][f"{self.screen_res[0]}x{self.screen_res[1]}"]["top_left"] |
| 111 | + bottom_right = self.values["fish_hooked_pos"][f"{self.screen_res[0]}x{self.screen_res[1]}"]["bottom_right"] |
| 112 | + bbox = (top_left[0], top_left[1], bottom_right[0], bottom_right[1]) |
| 113 | + screen_load = ImageGrab.grab(bbox=bbox).load() |
| 114 | + |
| 115 | + pixel = screen_load[0, 0] |
| 116 | + return self.color_distance(pixel, self.values["fish_hooked_blue"]) <= 10 |
| 117 | + |
| 118 | + def full_keepnet(self) -> bool: |
| 119 | + """ Check if the keepnet is full, return True if the keepnet is full, False otherwise. """ |
| 120 | + |
| 121 | + # grab the part of the screen where the keepnet is displayed |
| 122 | + top_left = self.values["full_keepnet_pos"][f"{self.screen_res[0]}x{self.screen_res[1]}"]["top_left"] |
| 123 | + bottom_right = self.values["full_keepnet_pos"][f"{self.screen_res[0]}x{self.screen_res[1]}"]["bottom_right"] |
| 124 | + bbox = (top_left[0], top_left[1], bottom_right[0], bottom_right[1]) |
| 125 | + screen_load = ImageGrab.grab(bbox=bbox).load() |
| 126 | + |
| 127 | + # check if any pixel in the keepnet is orange |
| 128 | + for i in range(0, np.shape(screen_load)[0]): |
| 129 | + if self.color_distance(screen_load[0, i], self.values["full_keepnet_orange"]) <= 10: |
| 130 | + return True |
| 131 | + return False |
| 132 | + |
| 133 | + def close_popups(self) -> bool: |
| 134 | + """ Close the pop-ups that might appear on the screen. Return True if any pop-up was closed, False otherwise. """ |
| 135 | + |
| 136 | + ret_changes = False |
| 137 | + |
| 138 | + changes = True |
| 139 | + while changes: |
| 140 | + changes = False |
| 141 | + |
| 142 | + # deal with normal pop-ups (achievements, extend license, etc.) |
| 143 | + for image in self.cv_templates["popups"]: |
| 144 | + inf = cv2.minMaxLoc(cv2.matchTemplate(cv2.cvtColor(np.array(ImageGrab.grab()), cv2.COLOR_RGB2GRAY), image, cv2.TM_SQDIFF)) |
| 145 | + if inf[0] <= 1000000: |
| 146 | + mouse.move(inf[2][0], inf[2][1], absolute=True, duration=0) |
| 147 | + mouse.click(button="left") |
| 148 | + changes = True |
| 149 | + ret_changes = True |
| 150 | + time.sleep(2) |
| 151 | + break |
| 152 | + if changes: |
| 153 | + continue |
| 154 | + |
| 155 | + # deal with buy popup (must be closed by clicking on the "X" button) |
| 156 | + screenshot = cv2.cvtColor(np.array(ImageGrab.grab()), cv2.COLOR_RGB2GRAY) |
| 157 | + inf = cv2.minMaxLoc(cv2.matchTemplate(screenshot, self.cv_templates["offers"]["buy"], cv2.TM_SQDIFF)) |
| 158 | + if inf[0] <= 1_000_000: |
| 159 | + inf2 = cv2.minMaxLoc(cv2.matchTemplate(screenshot, self.cv_templates["offers"]["x"], cv2.TM_SQDIFF)) |
| 160 | + shape = np.shape(self.cv_templates["offers"]["x"]) |
| 161 | + mouse.move(inf2[2][0] + shape[1] // 2, inf2[2][1] + shape[0] // 2, absolute=True, duration=0) |
| 162 | + mouse.click(button="left") |
| 163 | + changes = True |
| 164 | + ret_changes = True |
| 165 | + time.sleep(2) |
| 166 | + |
| 167 | + return ret_changes |
0 commit comments