Question/Issue:
I created my first Edge Impulse image recognition project: pictures were taken by my smartphone and immediately sent to Edge Impulse (3 categories). Finally Edge Impulse created a library for my Arduino nano 33 sense rev2. I downloaded and unzipped it to install it as an Arduino library. Then I loaded Arduino example ‘ei_nano_ble33_sense_camera.ino’, compiled it without any issues, and uploaded it to my Arduino nano 33 ble sense rev2 with an OV7670 camera attached.
I verified that the camera is working using another sketch.
However, no objects are detected at all, no matter what I try.
The program runs without error messages, and every few seconds this is printed on the serial monitor:
"
Starting inferencing in 2 seconds…
Taking photo…
Predictions (DSP: 3 ms., Classification: 217 ms., Anomaly: 0 ms.):
Object detection bounding boxes:
"
In Edge Impulse, under ‘create impulse’, I followed recommendation to put 48 x 48.
The camera is VGA (640 x 480).
In the Arduino sketch, this is defined (and I didn’t change it):
/* Constant variables ------------------------------------------------------- */
#define EI_CAMERA_RAW_FRAME_BUFFER_COLS 160
#define EI_CAMERA_RAW_FRAME_BUFFER_ROWS 120
I’m a novice Edge Impulse user; and I don’t know where to start looking to identify the problem. Any help would be much appreciated, and thank you in advance !
Project ID:
538600
Reproducibility:
- Always
Environment:
- Arduino nano 33 BLE sense rev2
- Build Environment Details: Arduino IDE 2.3.3
- OS Version: Windows 10
- Edge Impulse Version (Firmware): v1.29.0
Additional Information:
If anything is missing or you need extra info, please let me know.
Thank you