Skutt Kiln LCD Computer Vision Reader

A Python-based Computer Vision (CV) tool designed to monitor the state of a Skutt kiln’s segmented LCD display using a low-cost IP or RTSP camera feed. This script:

Captures frames from a Tapo or similar RTSP camera.
Uses an ArUco marker to establish a homography and rectify the image.
Applies feature-based matching to further refine the image alignment.
Composites images and uses contours to revise character origins.
Identifies active segments of the kiln’s LCD display.
Derives the kiln’s current temperature, firing state, and the displayed time.
Publishes these values to MQTT topic for home-assistant/HomeKit integration

Features

Captures images from an RTSP camera feed.
Uses ArUco markers for homography-based image rectification.
Employs template matching and feature detection for fine-tuned alignment.
Reads segmented digital displays to extract temperature, time, and kiln state.
Publishes extracted data to an MQTT broker for integration into home automation systems.
Supports flexible debugging options with image outputs at various processing stages.

License: BSD-3-Clause

Features

Robust Alignment:
Utilizes ArUco fiducials and feature-based matching (AKAZE) to correct for perspective distortions and lens issues.
Segmented LCD Detection:
Analyzes which LCD segments are active, applying logical rules and various thresholding methods (confidence intervals, Otsu-like approaches) to handle challenging layouts.
MQTT Integration:
Publishes temperature, time, and firing state readings to MQTT topics, simplifying integration with home automation frameworks.
Inexpensive: Utilizes sub-$20 commodity RTSP web camera

Tapo (TP-Link) C 110

Setup & Requirements

Camera Setup:
Position a webcam or RTSP-capable IP camera close to the kiln’s LCD and the attached ArUco marker (ID=8) 2"x2". A stable, front-facing view ensures accurate homography and segment detection.

Dependencies:

Tested with Python 3.12 conda

conda create --name skutt-monitor python=3.12
conda activate skutt-monitor
conda install -c conda-forge opencv
conda install -c conda-forge ffmpeg
conda install -c conda-forge paho-mqtt
conda install -c conda-forge python-dotenv

Installation

Clone the repository:

git clone https://github.com/rogerlew/skutt_kiln_cv
cd skutt_kiln_cv

Configure the .env file with your credentials and settings:

TAPO_USERNAME=<username>
PASSWORD=<password>
RTSP_IP=<camera-ip>
MQTT_BROKER=<broker-ip>
MQTT_PORT=<broker-port>
MQTT_USERNAME=<mqtt-username>
MQTT_PASSWORD=<mqtt-password>

Optionally create ramdisk at /ramdisk to write image files. script sets this to the working directory if it exists. Otherwise it creates a working dir relative to current directory.

Usage

Run the script from the command line:

conda run -n skutt-monitor python /workdir/skutt-kiln-monitor/skutt_read.py --no_debug

or to suppress debug

conda run -n skutt-monitor python /workdir/skutt-kiln-monitor/skutt_read.py --no_debug

Deployment

Create /var/log/skutt-monitor folder with permissions to write
add crontab

*/10 * * * *  /home/roger/miniconda3/bin/conda run -n skutt-monitor python /workdir/skutt-kiln-monitor/skutt_read.py --no_debug

Debugging & Customization

The script supports three levels of debugging:

Level 0: No debug information.
Level 1: Console output.
Level 2: Image outputs at key processing stages (e.g., grayscale conversion, rectification, character segmentation).

Fine tune bounding box for LCD Segment ROI and char_origins:

cropped = corrected_box[653:653+250, 664:664+565]

Update Template File

The template file (templates/templates.jpg) may need to be updated for the fine homography adjustment. The template file should have the LCD Display as rectilinear as possible. Web cameras with large fields of view have a great deal of lens distortion. To create the template image the camera was positioned 16" from the LCD screen such that the lens was centered vertically and horizontally with the screen. The image was captured and cropped to exclude background elements.

Home Assistant Integration

sensor:
  - platform: mqtt
    name: "Kiln Temperature"
    state_topic: "home/kiln/temperature"
    unit_of_measurement: "°F"
    value_template: "{{ value }}"
    device_class: "temperature"
    unique_id: "kiln_temperature_sensor"

For HomeKit bridge add to sensor.kiln_temperature to include_entities list

Physical Setup

License

This project is licensed under the BSD 3-Clause License. See the LICENSE file for details.

Visual Documentation

Computer Vision Processing

Grayscale Conversion

ArUco Marker Detection

Image Rectification

Template Feature Mapping

Image Rectification (2-pass)

Precise Contour Based Character Detection

The frames are composited using lighten blend.

Then contours are identified and the intersecting bounding boxes are used to revise the character origins

Character Segmentation

Original point based segmentation

Revised mask based segmentation

This is the template file used to define the segment masks

LCD Display ROI for reading

Example Output

...
Captured 7 frames


Processing frame 0...
  Detecting markers...
  Homography identification successful
  Rectifying image...
  Grayscale conversion,  Raw Min Value: 0.0, Raw Max Value: 225.216
  Detecting features for fine homography...
  Found 219 good matches
  Homography identification successful
  Corrected image saved as frame_0001,05_corrected.jpg

Processing frame 1...
  Rectifying image...
  Grayscale conversion,  Raw Min Value: 0.0, Raw Max Value: 225.216
  Corrected image saved as frame_0002,05_corrected.jpg

Processing frame 2...
  Rectifying image...
  Grayscale conversion,  Raw Min Value: 0.0, Raw Max Value: 225.216
  Corrected image saved as frame_0003,05_corrected.jpg

Processing frame 3...
  Rectifying image...
  Grayscale conversion,  Raw Min Value: 0.0, Raw Max Value: 225.216
  Corrected image saved as frame_0004,05_corrected.jpg

Processing frame 4...
  Rectifying image...
  Grayscale conversion,  Raw Min Value: 0.0, Raw Max Value: 225.216
  Corrected image saved as frame_0005,05_corrected.jpg

Processing frame 5...
  Rectifying image...
  Grayscale conversion,  Raw Min Value: 0.0, Raw Max Value: 225.216
  Corrected image saved as frame_0006,05_corrected.jpg

Processing frame 6...
  Rectifying image...
  Grayscale conversion,  Raw Min Value: 0.0, Raw Max Value: 225.216
  Corrected image saved as frame_0007,05_corrected.jpg

Revising character origin for character 1 at (44, 50) to (35, 47)
Revising character origin for character 2 at (170, 50) to (165, 47)
Revising character origin for character 3 at (299, 50) to (295, 47)
Revising character origin for character 4 at (426, 50) to (424, 47)
Reading processed frames...
  Reading frame 1...
Reading char 1 @ (35, 47)
read_segments (35, 47)
  Background Points [46, 52, 62, 61, 60, 49, 55, 57, 56, 54, 54, 53, 49, 64]
  Segment Points: [63, 68, 62, 64, 65, 67, 66, 64, 67, 66, 61, 64, 65, 64, 60]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 2 @ (165, 47)
read_segments (165, 47)
  Background Points [57, 57, 59, 60, 55, 51, 60, 60, 59, 54, 57, 58, 46, 60]
  Segment Points: [55, 58, 58, 61, 60, 61, 60, 56, 58, 60, 61, 60, 54, 62, 54]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 3 @ (295, 47)
read_segments (295, 47)
  Background Points [50, 51, 41, 47, 52, 54, 54, 51, 54, 54, 54, 51, 61, 54]
  Segment Points: [156, 154, 161, 54, 58, 58, 59, 58, 60, 59, 56, 59, 58, 54, 51]
  Otsu Point Segmentation
    Threshold: 153
    Segments: [1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 7
    Match Error: 0
Reading char 4 @ (424, 47)
read_segments (424, 47)
  Background Points [39, 52, 53, 36, 50, 39, 39, 45, 46, 50, 54, 50, 51, 46]
  Segment Points: [49, 148, 154, 55, 59, 52, 53, 52, 48, 55, 49, 53, 51, 52, 50]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 1
    Match Error: 0
  Display:   71

  Reading frame 2...
Reading char 1 @ (35, 47)
read_segments (35, 47)
  Background Points [60, 53, 56, 44, 56, 54, 58, 50, 56, 59, 46, 55, 58, 62]
  Segment Points: [67, 67, 66, 65, 67, 68, 67, 67, 69, 65, 66, 65, 65, 61, 61]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 2 @ (165, 47)
read_segments (165, 47)
  Background Points [55, 60, 42, 61, 54, 60, 53, 56, 50, 57, 54, 58, 58, 60]
  Segment Points: [158, 152, 60, 148, 163, 65, 179, 169, 60, 61, 64, 66, 59, 64, 156]
  Otsu Point Segmentation
    Threshold: 147
    Segments: [1, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1]
    Best Match: 2
    Match Error: 0
Reading char 3 @ (295, 47)
read_segments (295, 47)
  Background Points [51, 59, 52, 49, 54, 48, 56, 55, 54, 56, 42, 61, 54, 55]
  Segment Points: [53, 153, 156, 56, 58, 157, 179, 176, 58, 60, 58, 60, 58, 55, 56]
  Otsu Point Segmentation
    Threshold: 152
    Segments: [0, 1, 1, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 4
    Match Error: 0
Reading char 4 @ (424, 47)
read_segments (424, 47)
  Background Points [52, 39, 36, 44, 49, 43, 52, 40, 44, 51, 50, 48, 41, 51]
  Segment Points: [146, 149, 151, 145, 152, 155, 171, 162, 51, 58, 52, 59, 51, 56, 52]
  Otsu Point Segmentation
    Threshold: 144
    Segments: [1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 8
    Match Error: 0
  Display:  2.48

  Reading frame 3...
Reading char 1 @ (35, 47)
read_segments (35, 47)
  Background Points [50, 57, 58, 57, 62, 57, 57, 60, 57, 63, 57, 56, 58, 61]
  Segment Points: [64, 68, 64, 64, 69, 67, 66, 68, 67, 66, 64, 65, 67, 62, 60]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 2 @ (165, 47)
read_segments (165, 47)
  Background Points [38, 58, 40, 61, 37, 52, 57, 62, 40, 59, 57, 55, 58, 58]
  Segment Points: [156, 152, 59, 148, 163, 65, 175, 167, 60, 62, 64, 65, 61, 65, 157]
  Otsu Point Segmentation
    Threshold: 147
    Segments: [1, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1]
    Best Match: 2
    Match Error: 0
Reading char 3 @ (295, 47)
read_segments (295, 47)
  Background Points [52, 47, 50, 51, 53, 53, 54, 51, 58, 57, 47, 51, 50, 56]
  Segment Points: [55, 151, 154, 55, 61, 154, 182, 172, 59, 61, 56, 58, 54, 57, 52]
  Otsu Point Segmentation
    Threshold: 150
    Segments: [0, 1, 1, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 4
    Match Error: 0
Reading char 4 @ (424, 47)
read_segments (424, 47)
  Background Points [51, 59, 48, 47, 36, 38, 41, 51, 49, 37, 48, 49, 50, 50]
  Segment Points: [148, 149, 149, 146, 156, 156, 173, 163, 50, 60, 53, 58, 51, 57, 50]
  Otsu Point Segmentation
    Threshold: 145
    Segments: [1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 8
    Match Error: 0
  Display:  2.48

  Reading frame 4...
Reading char 1 @ (35, 47)
read_segments (35, 47)
  Background Points [49, 62, 59, 51, 52, 61, 54, 56, 59, 56, 50, 55, 60, 58]
  Segment Points: [154, 64, 63, 147, 154, 159, 66, 66, 67, 65, 65, 68, 67, 62, 58]
  Otsu Point Segmentation
    Threshold: 146
    Segments: [1, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: C
    Match Error: 0
Reading char 2 @ (165, 47)
read_segments (165, 47)
  Background Points [60, 43, 44, 52, 61, 61, 62, 59, 37, 56, 58, 64, 59, 57]
  Segment Points: [156, 152, 61, 58, 162, 162, 178, 174, 61, 63, 65, 66, 61, 67, 50]
  Otsu Point Segmentation
    Threshold: 151
    Segments: [1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
    Best Match: P
    Match Error: 0
Reading char 3 @ (295, 47)
read_segments (295, 47)
  Background Points [53, 58, 53, 56, 49, 47, 51, 52, 52, 51, 50, 51, 41, 48]
  Segment Points: [58, 54, 55, 149, 154, 155, 60, 53, 60, 60, 59, 60, 54, 52, 50]
  Otsu Point Segmentation
    Threshold: 148
    Segments: [0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: L
    Match Error: 0
Reading char 4 @ (424, 47)
read_segments (424, 47)
  Background Points [48, 48, 42, 48, 46, 48, 48, 48, 45, 42, 46, 48, 36, 45]
  Segment Points: [150, 52, 53, 56, 57, 53, 50, 52, 161, 162, 48, 53, 49, 51, 51]
  Otsu Point Segmentation
    Threshold: 149
    Segments: [1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0]
    Best Match: T
    Match Error: 0
  Display: CPLT

  Reading frame 5...
Reading char 1 @ (35, 47)
read_segments (35, 47)
  Background Points [68, 64, 54, 57, 57, 54, 58, 61, 50, 59, 52, 61, 69, 65]
  Segment Points: [154, 66, 65, 148, 157, 159, 67, 65, 68, 64, 68, 66, 66, 63, 61]
  Otsu Point Segmentation
    Threshold: 147
    Segments: [1, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: C
    Match Error: 0
Reading char 2 @ (165, 47)
read_segments (165, 47)
  Background Points [43, 57, 57, 61, 52, 65, 43, 50, 61, 43, 63, 51, 41, 50]
  Segment Points: [156, 151, 58, 59, 164, 161, 179, 167, 62, 62, 66, 65, 66, 64, 51]
  Otsu Point Segmentation
    Threshold: 150
    Segments: [1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
    Best Match: P
    Match Error: 0
Reading char 3 @ (295, 47)
read_segments (295, 47)
  Background Points [56, 51, 53, 53, 50, 60, 49, 58, 61, 50, 53, 53, 52, 45]
  Segment Points: [58, 56, 55, 151, 150, 156, 60, 54, 59, 60, 59, 60, 52, 53, 50]
  Otsu Point Segmentation
    Threshold: 149
    Segments: [0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: L
    Match Error: 0
Reading char 4 @ (424, 47)
read_segments (424, 47)
  Background Points [47, 44, 44, 44, 51, 42, 51, 50, 48, 40, 47, 47, 43, 50]
  Segment Points: [150, 53, 54, 57, 56, 53, 50, 53, 160, 167, 50, 53, 52, 52, 49]
  Otsu Point Segmentation
    Threshold: 149
    Segments: [1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0]
    Best Match: T
    Match Error: 0
  Display: CPLT

  Reading frame 6...
Reading char 1 @ (35, 47)
read_segments (35, 47)
  Background Points [59, 60, 64, 62, 61, 58, 54, 56, 59, 59, 62, 58, 54, 55]
  Segment Points: [62, 66, 62, 63, 68, 65, 64, 65, 68, 65, 62, 65, 65, 61, 58]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 2 @ (165, 47)
read_segments (165, 47)
  Background Points [51, 62, 56, 57, 49, 56, 55, 54, 55, 52, 38, 57, 53, 57]
  Segment Points: [54, 57, 58, 61, 62, 64, 59, 58, 62, 60, 63, 62, 56, 64, 51]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 3 @ (295, 47)
read_segments (295, 47)
  Background Points [52, 55, 53, 56, 55, 58, 45, 54, 40, 48, 60, 48, 52, 46]
  Segment Points: [155, 152, 156, 55, 58, 59, 59, 57, 60, 62, 55, 57, 58, 56, 52]
  Otsu Point Segmentation
    Threshold: 151
    Segments: [1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 7
    Match Error: 0
Reading char 4 @ (424, 47)
read_segments (424, 47)
  Background Points [48, 49, 49, 39, 39, 49, 46, 36, 51, 37, 56, 50, 50, 51]
  Segment Points: [49, 152, 153, 57, 57, 54, 51, 52, 48, 56, 46, 55, 50, 55, 52]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 1
    Match Error: 0
  Display:   71

  Reading frame 7...
Reading char 1 @ (35, 47)
read_segments (35, 47)
  Background Points [48, 62, 58, 59, 51, 57, 60, 57, 57, 56, 53, 62, 54, 54]
  Segment Points: [64, 66, 65, 64, 67, 66, 66, 67, 68, 65, 63, 64, 65, 64, 61]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 2 @ (165, 47)
read_segments (165, 47)
  Background Points [50, 48, 59, 51, 40, 60, 45, 60, 45, 45, 54, 46, 40, 60]
  Segment Points: [57, 59, 59, 60, 63, 65, 60, 58, 58, 61, 63, 60, 56, 63, 50]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match:
    Match Error: 0
Reading char 3 @ (295, 47)
read_segments (295, 47)
  Background Points [53, 58, 56, 55, 57, 53, 39, 60, 48, 46, 56, 47, 49, 47]
  Segment Points: [156, 157, 156, 54, 59, 61, 58, 58, 61, 60, 57, 56, 61, 56, 51]
  Otsu Point Segmentation
    Threshold: 155
    Segments: [1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 7
    Match Error: 0
Reading char 4 @ (424, 47)
read_segments (424, 47)
  Background Points [42, 52, 27, 42, 44, 39, 39, 36, 36, 42, 49, 45, 50, 37]
  Segment Points: [49, 153, 151, 56, 58, 51, 52, 53, 48, 57, 49, 54, 51, 54, 51]
  Otsu Point Segmentation
    Threshold: 100
    Segments: [0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
    Best Match: 1
    Match Error: 0
  Display:   71

Frame Semantics:
  Frame 1:   71
    Unrecognized:   71
  Frame 2:  2.48
    Time: 02:48
  Frame 3:  2.48
    Time: 02:48
  Frame 4: CPLT
    State: Complete
  Frame 5: CPLT
    State: Complete
  Frame 6:   71
    Temperature: 71.0
  Frame 7:   71
    Unrecognized:   71
Published temperature: 71.0 °F to topic: home/kiln/temperature
Published time: 02:48 to topic: home/kiln/time
Published state: Complete to topic: home/kiln/state
state=Complete, temp=71.0, time=02:48

Name		Name	Last commit message	Last commit date
Latest commit History 31 Commits
templates		templates
.gitignore		.gitignore
LICENSE.md		LICENSE.md
README.md		README.md
skutt_read.py		skutt_read.py

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

Skutt Kiln LCD Computer Vision Reader

Features

Features

Setup & Requirements

Installation

Usage

Deployment

Debugging & Customization

Physical Setup

License

Visual Documentation

Computer Vision Processing

Grayscale Conversion

ArUco Marker Detection

Image Rectification

Template Feature Mapping

Image Rectification (2-pass)

Precise Contour Based Character Detection

Character Segmentation

LCD Display ROI for reading

Example Output

About

Releases

Packages

Contributors 2

Languages

License

rogerlew/skutt-kiln-monitor

Folders and files

Latest commit

History

Repository files navigation

Skutt Kiln LCD Computer Vision Reader

Features

Features

Setup & Requirements

Installation

Usage

Deployment

Debugging & Customization

Physical Setup

License

Visual Documentation

Computer Vision Processing

Grayscale Conversion

ArUco Marker Detection

Image Rectification

Template Feature Mapping

Image Rectification (2-pass)

Precise Contour Based Character Detection

Character Segmentation

LCD Display ROI for reading

Example Output

About

Resources

License

Stars

Watchers

Forks

Releases

Packages 0

Contributors 2

Languages

Packages