Image Classification with Edge AI
Classifying entire images into predefined categories without localization. The model outputs a single class label per frame — no bounding boxes or object positions. Uses quantized MobileNet, EfficientNet-Lite, or custom CNN architectures optimized for microcontrollers. Simpler than object detection with lower resource requirements. Common applications include quality inspection, scene recognition, and presence detection.
Hardware Requirements
| Minimum RAM | 128 KB |
| Minimum Flash | 512 KB |
| Sensor Inputs | camera |
| Typical Model Size | 150 KB (quantized int8) |
| Minimum Clock | 80 MHz |
Compatible Microcontrollers
ESP32
Espressif
520 KB RAM · 240 MHz
$5–$15 (dev board)
ESP32-C3
Espressif
400 KB RAM · 160 MHz
$4–$10 (dev board)
ESP32-C6
Espressif
512 KB RAM · 160 MHz
$5–$15 (dev board)
ESP32-S3
Espressif
512 KB RAM · 240 MHz
$10–$25 (dev board)
i.MX RT1062
NXP
1024 KB RAM · 600 MHz
$25–$40 (dev board)
nRF52840
Nordic Semiconductor
256 KB RAM · 64 MHz
$20–$35 (dev board)
RA6M5
Renesas
512 KB RAM · 200 MHz
$25–$50 (dev board)
STM32F4
STMicroelectronics
192 KB RAM · 168 MHz
$10–$30 (dev board)
STM32F7
STMicroelectronics
512 KB RAM · 216 MHz
$25–$60 (dev board)
STM32H7
STMicroelectronics
1024 KB RAM · 480 MHz
$30–$80 (dev board)
STM32U5
STMicroelectronics
786 KB RAM · 160 MHz
$20–$50 (dev board)
Hardware Guides
ESP32-C3 Image Classification with Edge Impulse
Espressif's ESP32-C3 is a solid choice for image classification using Edge Impulse. The risc-v core at 160 MHz with 400 KB SRAM accommodates…
ESP32-C3 Image Classification with TFLite Micro
The ESP32-C3 handles image classification effectively with TFLite Micro. 400 KB SRAM at 160 MHz provides 3.1x headroom over the 128 KB requi…
ESP32-C6 Image Classification with Edge Impulse
The ESP32-C6 handles image classification effectively with Edge Impulse. 512 KB SRAM at 160 MHz provides 4.0x headroom over the 128 KB requi…
ESP32-C6 Image Classification with TFLite Micro
Running image classification on the ESP32-C6 with TFLite Micro is practical. 512 KB SRAM meets the 128 KB minimum with 4.0x headroom. The 16…
ESP32 Image Classification with Edge Impulse
The ESP32 handles image classification effectively with Edge Impulse. 520 KB SRAM at 240 MHz provides 4.1x headroom over the 128 KB requirem…
ESP32 Image Classification with TFLite Micro
Running image classification on the ESP32 with TFLite Micro is practical. 520 KB SRAM meets the 128 KB minimum with 4.1x headroom. The 240 M…
ESP32-S3 Image Classification with Edge Impulse
The ESP32-S3 is an excellent match for image classification with Edge Impulse. 512 KB SRAM delivers 4.0x the 128 KB minimum while 240 MHz pr…
ESP32-S3 Image Classification with TFLite Micro
The ESP32-S3 is an excellent match for image classification with TFLite Micro. 512 KB SRAM delivers 4.0x the 128 KB minimum while 240 MHz pr…
i.MX RT1062 Image Classification with CMSIS-NN
For image classification, the i.MX RT1062 with CMSIS-NN scores Excellent. Its 1024 KB internal SRAM (8.0x the required 128 KB) and 600 MHz c…
i.MX RT1062 Image Classification with TFLite Micro
The i.MX RT1062 is an excellent match for image classification with TFLite Micro. 1024 KB SRAM delivers 8.0x the 128 KB minimum while 600 MH…
nRF52840 Image Classification with Edge Impulse
The nRF52840 handles image classification effectively with Edge Impulse. 256 KB SRAM at 64 MHz provides 2.0x headroom over the 128 KB requir…
nRF52840 Image Classification with TFLite Micro
Running image classification on the nRF52840 with TFLite Micro is practical. 256 KB SRAM meets the 128 KB minimum with 2.0x headroom. The 64…
RA6M5 Image Classification with CMSIS-NN
Running image classification on the RA6M5 with CMSIS-NN is practical. 512 KB SRAM meets the 128 KB minimum with 4.0x headroom. The 200 MHz c…
RA6M5 Image Classification with TFLite Micro
Renesas's RA6M5 is a solid choice for image classification using TFLite Micro. The cortex-m33 core at 200 MHz with 512 KB SRAM accommodates …
STM32F4 Image Classification with Edge Impulse
The STM32F4 handles image classification effectively with Edge Impulse. 192 KB SRAM at 168 MHz provides 1.5x headroom over the 128 KB requir…
STM32F4 Image Classification with TFLite Micro
The STM32F4 handles image classification effectively with TFLite Micro. 192 KB SRAM at 168 MHz provides 1.5x headroom over the 128 KB requir…
STM32F7 Image Classification with CMSIS-NN
STMicroelectronics's STM32F7 excels at image classification via CMSIS-NN. The 1-core cortex-m7 at 216 MHz with 512 KB SRAM handles 150 KB qu…
STM32F7 Image Classification with TFLite Micro
For image classification, the STM32F7 with TFLite Micro scores Excellent. Its 512 KB internal SRAM (4.0x the required 128 KB) and 216 MHz cl…
STM32H7 Image Classification with CMSIS-NN
The STM32H7 is an excellent match for image classification with CMSIS-NN. 1024 KB SRAM delivers 8.0x the 128 KB minimum while 480 MHz proces…
STM32H7 Image Classification with TFLite Micro
The STM32H7 is an excellent match for image classification with TFLite Micro. 1024 KB SRAM delivers 8.0x the 128 KB minimum while 480 MHz pr…
STM32U5 Image Classification with CMSIS-NN
STMicroelectronics's STM32U5 is a solid choice for image classification using CMSIS-NN. The cortex-m33 core at 160 MHz with 786 KB SRAM acco…
STM32U5 Image Classification with TFLite Micro
STMicroelectronics's STM32U5 is a solid choice for image classification using TFLite Micro. The cortex-m33 core at 160 MHz with 786 KB SRAM …
Industry Applications
Orchestrate Image Classification with ForestHub
Your devices run image classification on-device. ForestHub on your Linux edge gateway ingests their results over MQTT/Modbus/OPC-UA, orchestrates the sense-reason-act loop as an auditable graph, and acts on the line — the LLM is one node among many.