[Python] 纯文本查看 复制代码import cv2
import numpy as np
# 读取图像
# image = cv2.imread("./test.png")
image = cv2.imread("./t2.jpg")
ball_color = ['yellow', 'pink'] # 需要识别的颜色 支持多个颜色
color_dist = {
'pink': {'Lower': np.array([150, 43, 46]), 'Upper': np.array([175, 255, 255])},
'yellow': {'Lower': np.array([21, 43, 46]), 'Upper': np.array([34, 255, 255])},
'purple': {'Lower': np.array([125, 43, 46]), 'Upper': np.array([155, 255, 255])}
}
gs_frame = cv2.GaussianBlur(image, (5, 5), 0) # 高斯模糊
hsv = cv2.cvtColor(gs_frame, cv2.COLOR_BGR2HSV) # 转化成HSV图像
erode_hsv = cv2.erode(hsv, None, iterations=2) # 腐蚀 粗的变细
colorArr = []
for colorValue in ball_color: #获取所有匹配的区域块
inRange_hsv = cv2.inRange(erode_hsv, color_dist[colorValue]['Lower'], color_dist[colorValue]['Upper'])
c = cv2.findContours(inRange_hsv.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[-2]
colorArr += c
mask = max(colorArr, key=cv2.contourArea) #获取最大色块
rect = cv2.minAreaRect(mask)
box = cv2.boxPoints(rect)
box = np.intp(box)
# 计算边界框的中心点坐标
center_x = int((box[0][0] + box[2][0]) / 2)
center_y = int((box[0][1] + box[2][1]) / 2)
# 随机生成20个点位
points = []
for _ in range(20):
x = np.random.randint(center_x - 8, center_x + 8)
y = np.random.randint(center_y - 8, center_y + 8)
points.append((x, y))
# 计算20个点位的平均颜色值
avg_color = np.zeros(3)
for point in points:
color = hsv[point[1], point[0]]
avg_color += color
avg_color = tuple(map(int, avg_color / len(points)))
# 打印平均颜色值
print("平均颜色值：", avg_color)
# 判断平均颜色值接近黄色还是接近紫色
yellow_dist = np.linalg.norm(avg_color - color_dist['yellow']['Lower'])
purple_dist = np.linalg.norm(avg_color - color_dist['purple']['Lower'])
# 接口应在这里返回结果
if yellow_dist
示例图片
示例结果
这个东西感觉并不准确 但是也没有找到好的方法，小白不怎么会写python。 想要把这个代码做成接口形式，试了试不知道为什么跑不起来。postman 传了图片但是断点查看并没有图片，不知道咋回事
