Hello. I have OpenMV H7 Plus, Pan-tilt servo and LCD Shield. I encounter problem when I running the code for object detection and tracking. Pan-tilt and lcd are connected on the same pin which P7 and P8. Thus, I faced a conflict pin means when Lcd turn on, the servo unable to move. Since I am using the example code from OpenMV IDE and in the example code there is no define of number pin. Let say I change pin LCD for I2C to pin 4 and pin 5, is it the code still same? Below is my code
import sensor, image, time, os, ml, math, uos, gc, display
from ulab import numpy as np
from pyb import Servo
from pid import PID
import pyb
# Initialize pan and tilt servos
pan_servo = Servo(1)
tilt_servo = Servo(2)
# Servo calibration
pan_servo.calibration(500, 2500, 500)
tilt_servo.calibration(500, 2500, 500)
# PID parameters for smooth tracking
pan_pid = PID(p=0.2, i=0, imax=90)
tilt_pid = PID(p=0.2, i=0, imax=90)
# Initialize camera
sensor.reset()
sensor.set_pixformat(sensor.RGB565)
sensor.set_framesize(sensor.QQVGA2) # Modify as you like.
sensor.set_vflip(True)
sensor.skip_frames(time=2000)
sensor.set_auto_whitebal(False)
# Load ML model
net = None
labels = None
min_confidence = 0.7 # Confidence threshold
try:
net = ml.Model("trained14.tflite", load_to_fb=uos.stat('trained14.tflite')[6] > (gc.mem_free() - (64*1024)))
except Exception as e:
raise Exception('Failed to load "trained14.tflite". Copy the .tflite and labels14.txt files to the device. (' + str(e) + ')')
try:
labels = [line.rstrip('\n') for line in open("labels14.txt")]
except Exception as e:
raise Exception('Failed to load "labels14.txt". Copy the .tflite and labels14.txt files to the device. (' + str(e) + ')')
# Colors for different objects
colors = [(255, 0, 0), (0, 255, 0), (255, 255, 0), (0, 0, 255), (255, 0, 255), (0, 255, 255), (255, 255, 255)]
# Detection confidence threshold
threshold_list = [(math.ceil(min_confidence * 255), 255)]
def fomo_post_process(model, inputs, outputs):
"""Post-processing for FOMO object detection."""
ob, oh, ow, oc = model.output_shape[0]
x_scale = inputs[0].roi[2] / ow
y_scale = inputs[0].roi[3] / oh
scale = min(x_scale, y_scale)
x_offset = ((inputs[0].roi[2] - (ow * scale)) / 2) + inputs[0].roi[0]
y_offset = ((inputs[0].roi[3] - (ow * scale)) / 2) + inputs[0].roi[1]
detections = [[] for _ in range(oc)]
for i in range(oc):
img = image.Image(outputs[0][0, :, :, i] * 255)
blobs = img.find_blobs(threshold_list, x_stride=1, y_stride=1, area_threshold=1, pixels_threshold=1)
for b in blobs:
x, y, w, h = b.rect()
score = img.get_statistics(thresholds=threshold_list, roi=b.rect()).l_mean() / 255.0
x, y, w, h = int(x * scale + x_offset), int(y * scale + y_offset), int(w * scale), int(h * scale)
detections[i].append((x, y, w, h, score))
return detections
clock = time.clock()
lcd = display.SPIDisplay()
while True:
lcd.write(sensor.snapshot()) # Take a picture and display the image.
clock.tick()
img = sensor.snapshot()
largest_obj = None
largest_area = 0
detected_label = None
# Process object detection
for i, detection_list in enumerate(net.predict([img], callback=fomo_post_process)):
if i == 0 or len(detection_list) == 0: # Ignore background class or empty detections
continue
for x, y, w, h, score in detection_list:
if score < min_confidence:
continue # Ignore low-confidence detections
area = w * h
if area > largest_area:
largest_area = area
largest_obj = (x, y, w, h)
detected_label = labels[i] if i < len(labels) else "Unknown"
if largest_obj:
x, y, w, h = largest_obj
centroid_x = x + w // 2
centroid_y = y + h // 2
# Draw tracking dot only
img.draw_circle(centroid_x, centroid_y, 6, color=colors[i % len(colors)])
# Center calculations
img_center_x, img_center_y = img.width() / 2, img.height() / 2
# Error calculation
pan_error = centroid_x - img_center_x
tilt_error = centroid_y - img_center_y
# PID control
pan_output = pan_pid.get_pid(pan_error, 1) / 2
tilt_output = tilt_pid.get_pid(tilt_error, 1) /2
print("pan_output:", pan_output)
# Invert the pan PID output correctly
pan_servo.angle(pan_servo.angle() + pan_output)
# For tilt, the direction remains as originally coded.
tilt_servo.angle(tilt_servo.angle() + tilt_output)
print(f"FPS: {clock.fps()}")
