Image-Filtering-in-Assembly--RISC-V-

This project implements image processing in Assembly (RISC-V) by applying a convolution filter to a grayscale image.

Original Image

Filtered-image-assembly

Filtered-image-python

The image is represented as a byte array (.byte), and the program iterates over this matrix applying a kernel (mask) to generate a processed output image.

⚙️ Technologies Used

Language: RISC-V Assembly Execution environment: RARS (or similar simulator)

🧠 Concepts Covered

Memory manipulation (.data and .text) Linearized matrix addressing Loops in Assembly Register usage Macro definition (.macro) Digital image processing Convolution filtering

🖼️ Image Structure

The image is stored as a linear array: imagem: .byte ... Its dimensions are defined as:

linhas: .word 119 colunas: .word 175 control: .word 176 # row width in memory 🔍 Algorithm Overview

The program iterates pixel by pixel (ignoring borders) and applies a convolution filter.

🧩 Kernel (Mask) The filter used is:

-1 -2 -1 0 0 0 1 2 1

This kernel is similar to the vertical Sobel operator, commonly used for edge detection.

⚙️ Processing Steps

Read the central pixel Apply convolution Multiply neighboring pixels by kernel weights Accumulate the results Normalize the result

If < 0 → set to 0 If ≥ 256 → set to 255

Store the result Save into the secondary matrix Print the value Output to console

🔁 Macro Used

To reduce code repetition, the following macro is defined:

.macro mascaraS lb s4,0(s10) mul s5,s4,t6 add s9,s9,s5 .end_macro 📌 Purpose:

Loads a neighboring pixel Multiplies it by the kernel weight Adds it to the accumulator

💾 Output

The processed values are: Printed to the console Stored in memory (matrizsecundaria2)

matrizsecundaria2: .space 21120

▶️ How to Run

Open the code in a RISC-V simulator (e.g., RARS)

Assemble the program Run it Observe the output in the console

📊 Expected Result

The program generates a new image with enhanced vertical edges, useful for: Computer vision Image processing Edge detection

🚀 Possible Improvements

Implement additional filters (horizontal Sobel, Laplacian, blur) Display the image graphically Handle border pixels properly Optimize register usage Further modularize the code

👨‍💻 Author

This project was developed as practice for: Low-level programming Memory manipulation Image processing in Assembly